{"version":1,"narrative":{"title":"Emre’s Tesseract","intro":"Welcome in. From the tone of a Turkish sentence to a quantum measurement, these are the questions I keep coming back to.","domains":[{"id":"meaning","title":"Meaning and ontologies","entries":["project:meaning-native-computing","note:ontologies-and-data-structures","note:category-theory","patent:2026-007907","patent:2026-006289","patent:2025-008786"]},{"id":"language","title":"Natural language processing","entries":["repository:turkish-binary-reviews","work:turkish-asr","topic:language","work:chat-assistant","work:ai-pos","work:qwen-grpo","work:turkish-reasoning-benchmark","work:multilingual-chatbots","work:ad-text-classification","work:weight-editing","work:bloom","patent:2026-007909","patent:2023-011870","model:turkish-sentiment","model:whisper-tr","model:gemma-wiki-tr","model:java-method-vectors","model:gemma-reasoning-tr","model:abstract-title","model:xlsr-tr","model:qwen-grpo","repository:webcli","study:turkish-sentiment-evaluation"]},{"id":"learning","title":"Machine learning","entries":["repository:regression-atlas","work:weight-editing","work:coverage-deferral","patent:2025-011133","repository:fairness-metrics","repository:worldcup-nowcast","repository:turkish-binary-reviews","study:coverage-deferral"]},{"id":"confidence","title":"Conformal prediction","entries":["work:coverage-deferral","topic:uncertainty","topic:fairness","project:meaning-native-computing","work:quantum-conformal","work:tara","work:adversarial-quantum","patent:2026-004441","repository:regression-atlas","repository:bird-sql","repository:fairness-metrics","repository:worldcup-nowcast","study:coverage-deferral"]},{"id":"quantum","title":"Quantum computing","entries":["work:quantum-conformal","topic:quantum","note:quantum-measurement","work:tara","work:adversarial-quantum","patent:2026-004441","repository:quantum-clocks"]},{"id":"invention","title":"Inventions","entries":["patent:2026-004441","patent:2023-011901","topic:invention","work:chat-assistant","work:micro-robot-rescue","work:ai-pos","work:virtual-input","work:personalized-suggestions","patent:2026-007909","patent:2026-007907","patent:2026-006294","patent:2026-006289","patent:2026-004444","patent:2026-002909","patent:2026-001992","patent:2025-019989","patent:2025-011133","patent:2025-008786","patent:2025-008273","patent:2025-008055","patent:2025-008032","patent:2025-008028","patent:2025-007997","patent:2025-007995","patent:2024-003815","patent:2024-003813","patent:2024-003592","patent:2023-011870","patent:2023-011830","patent:2023-011601","patent:2023-011600","patent:2023-011599","patent:2023-004244","patent:2023-004047","patent:2022-016679","patent:2022-016395"]},{"id":"agents","title":"Agent systems","entries":["work:agent-chronicle","topic:agents","work:chat-assistant","work:micro-robot-rescue","work:weight-editing","patent:2025-007995","repository:bird-sql","repository:webcli","repository:paperbanana-test","repository:agent-chronicle"]}],"stories":[{"id":"project:meaning-native-computing","domain":"meaning","title":"A program’s next move","text":"In Meaning-Native Computing, I ask whether knowing a program’s rules helps a predictor anticipate its next instruction. An independent check decides whether that step can run.","year":"","companions":["note:ontologies-and-data-structures"],"url":"/?scene=project%3Ameaning-native-computing#journey"},{"id":"repository:turkish-binary-reviews","domain":"language","title":"The tone of a Turkish review","text":"I trained a Turkish BERT model to tell positive reviews from negative ones. In 2022 I released another that could also recognise a neutral review.","year":"2021","companions":["model:turkish-sentiment"],"url":"/?scene=repository%3Aturkish-binary-reviews#journey"},{"id":"work:turkish-asr","domain":"language","title":"Listening to Turkish","text":"We taught a speech model trained on 53 languages to transcribe Turkish using recordings from Common Voice. On recordings held back for testing, its word error rate was 23%.","year":"2024","companions":["model:xlsr-tr","model:whisper-tr"],"url":"/?scene=work%3Aturkish-asr#journey"},{"id":"repository:regression-atlas","domain":"learning","title":"How wide should a forecast be?","text":"A wide interval can make a forecast look reliable. Across 67 datasets, I compared how often each method caught the actual value and how wide its intervals had to be.","year":"2026","companions":["work:coverage-deferral","work:quantum-conformal"],"url":"/?scene=repository%3Aregression-atlas#journey"},{"id":"work:coverage-deferral","domain":"confidence","title":"Who waits for a second look?","text":"In my simulation on the Adult dataset, calibrating each group separately brought their coverage rates closer. Yet it widened the gap in how often their cases went to human review.","year":"2025","companions":["study:coverage-deferral","repository:fairness-metrics"],"url":"/?scene=work%3Acoverage-deferral#journey"},{"id":"work:quantum-conformal","domain":"quantum","title":"Two qubits, four outcomes","text":"Two qubits can give four different answers. I used earlier examples to estimate the uncertainty in all four probabilities together, then tested those estimates on fresh data.","year":"2025","companions":["patent:2026-004441","note:quantum-measurement"],"url":"/?scene=work%3Aquantum-conformal#journey"},{"id":"patent:2026-004441","domain":"invention","title":"Customer groups in noisy data","text":"How stable are the groups we find in financial data? In application TR 2026/004441, we propose using quantum clustering to form the groups and conformal prediction to quantify the uncertainty.","year":"2026","companions":["work:quantum-conformal"],"url":"/?scene=patent%3A2026-004441#journey"},{"id":"patent:2023-011901","domain":"invention","title":"A gap a finger wide","text":"Our Turkish filing describes a jointed micro-robot that could carry a camera and a tether into rubble. The same invention appeared in an international patent publication in 2025.","year":"2023","companions":["work:micro-robot-rescue"],"url":"/?scene=patent%3A2023-011901#journey"},{"id":"work:agent-chronicle","domain":"agents","title":"After an agent says “done”","text":"I built Agent Chronicle so I could retrace a coding session. From a decision I can follow the changes to the files and open the checks behind the result.","year":"2026","companions":["repository:agent-chronicle","patent:2025-007995"],"url":"/?scene=work%3Aagent-chronicle#journey"},{"id":"work:weight-editing","domain":"learning","title":"Can one model mend another?","text":"When I let language models edit a damaged classifier, one run raised balanced accuracy slightly while overall accuracy fell. On intact SmolLM2, the searches kept the original weights.","year":"2026","companions":["work:coverage-deferral"],"url":"/?scene=work%3Aweight-editing#journey"}]},"lang":"en","startId":"profile:emre","stops":["profile:emre","project:meaning-native-computing","topic:uncertainty","work:coverage-deferral","topic:fairness","topic:quantum","patent:2026-004441","topic:invention","topic:agents","work:weight-editing","topic:language"],"transitions":["In my independent research, I ask whether knowing a program’s rules helps predict its next instruction.","Before a predicted step takes effect, an independent check must confirm that the rules allow it.","When a model is unsure, it can pass a decision to a person for review.","I compare how often people in different groups have their cases sent for review.","Quantum research raises a different question: what exactly are we measuring?","Turkish application 2026/004441 proposes combining quantum entanglement and conformal prediction to group customers despite noisy data.","This is one of my 30 filings in Türkiye: 28 patent applications and two utility model applications.","Several of the inventions use software to coordinate actions across systems.","Here, language models propose changes to another model’s weights, and I measure what happens.","Other projects bring language models to Turkish tasks such as classifying reviews."],"labels":{"kinds":{"topic":"Topic","work":"Research work","patent":"Application","project":"Project","affiliation":"Affiliation","photo":"Photograph","section":"Collection","study":"Study","reader":"Reading view","model":"Model","repository":"Repository","note":"Research note","profile":"Profile","contact":"Contact"},"relations":{"about":"Research topic","related":"Related entry","contains":"Collection membership","documents":"Documentation","affiliation":"Institutional link","readableAs":"Reading view","sameProject":"Same project","family":"Same patent family","contact":"Public profile link"},"inverseRelations":{"about":"Research topic","related":"Related entry","contains":"Containing collection","documents":"Documentation","affiliation":"Institutional link","readableAs":"Reading view","sameProject":"Same project","family":"Same patent family","contact":"Public profile link"},"sections":{"portfolio":"Applications in Türkiye","lab":"Models & code","works":"Research and publications","notebook":"Four questions from my notebook","album":"My photo album","contact":"Contact & profiles","readers":"Index, graph and profile","selected":"Selected work","research":"Research areas"}},"nodes":[{"id":"profile:emre","kind":"profile","title":"Davut Emre Taşar","summary":"I’m an engineer and applied scientist on BBVA’s AI Innovation team, part of AI Transformation. I’m doing a PhD at the University of Navarra on fairness in AI.","details":["I’m from İzmir, on Turkey’s Aegean coast. I used to be a civil engineer. Then I took an arrow to the knee and traded my hard hat for a lab coat.","I’m an engineer and applied scientist on BBVA’s AI Innovation team. At the University of Navarra, I’m doing a PhD on fairness in AI. My interests range from what makes a good team to abstract maths, quantum physics and philosophy. I especially enjoy finding a link between two ideas that seem to have nothing in common.","Madrid is home now. I’m exploring the city, learning Spanish and getting to know the cultures of Spain and Latin America. At home I paint, play chess and make up little stories for my daughter."],"path":"/profile/","links":[],"image":"/assets/emre-espresso.jpg","meta":"Madrid","fields":[]},{"id":"section:research","kind":"section","title":"Research areas","summary":"Uncertainty · Fairness · Agent systems · Quantum · Invention · Language","details":["Can intelligence know its limits?","Who carries the uncertainty?","How can you check what an agent actually did?","What can a measurement tell us?","How small would a rescue robot need to be?","How do machines understand us?"],"path":"/#explore","links":[],"fields":[]},{"id":"section:selected","kind":"section","title":"Selected work","summary":"Can a computer predict which instructions will run next if it knows the rules of the program?","details":["How much can a finite number of measurements tell us?","What could reach someone through a gap only a finger wide?","Can one model edit the weights of another?"],"path":"/#selected-exhibits","links":[],"fields":[]},{"id":"section:portfolio","kind":"section","title":"Applications in Türkiye","summary":"Public registry and publication records name Emre Taşar as an inventor on 28 patent applications and 2 utility model applications in Türkiye. Filing numbers, titles and dates come from the TÜRKPATENT extract of 2026-08-31, with protection types from the inventor’s list. Current legal status is checked separately in the register. Some inventions also have an international publication.","details":["Applicant · TÜRKİYE GARANTİ BANKASI ANONİM ŞİRKETİ","Register snapshot · Aug 31, 2026","National filings may overlap with PCT publications. The total counts applications filed in Türkiye.","Portfolio-wide legal status is not provided by the dated extract. Individual grant decisions cite the national register."],"path":"/#patents","links":[{"label":"Official TÜRKPATENT register","url":"https://www.turkpatent.gov.tr/arastirma-yap?form=patent"}],"fields":[]},{"id":"section:lab","kind":"section","title":"Models & code","summary":"Model · 8   /   Repository · 8","details":["Turkish Sentiment Analysis · TSAM — I trained this BERT model to read the tone of a Turkish review. It scores the sentence for positive, neutral and negative sentiment.","Fairness Metrics — Give the Python library features, predictions, actual outcomes and a protected group. It returns a pandas table comparing quantities such as false-positive, false-negative and recall rates across groups.","Agent Chronicle — Read a Codex or Claude session as a timeline of decisions, changes and checks. Each recorded event gives the developer a way to inspect the evidence behind a progress report."],"path":"/#lab","links":[],"fields":[]},{"id":"section:works","kind":"section","title":"Research and publications","summary":"Search research, Turkish patent and utility model applications, models and code repositories in one place.","details":["The coverage–deferral trade-off — Whose case gets a second look?","Quantum-enhanced conformal methods — How much can a finite number of measurements tell us?","Editing neural networks with language models — Can one model edit the weights of another?"],"path":"/#work","links":[],"fields":[]},{"id":"section:notebook","kind":"section","title":"Four questions from my notebook","summary":"Ontologies and data structures · Meaning-Native Computing · Category theory · Quantum measurement","details":["What changes in a graph when a train runs late?","Can a computer predict which instructions will run next if it knows the rules of the program?","What survives when I compose two transformations?","What if I ask the same qubit a different question?"],"path":"/#idea-notebook-title","links":[],"fields":[]},{"id":"section:album","kind":"section","title":"My photo album","summary":"Emre beside an Elektra espresso machine · Emre presenting his PhD research at Navarra · Emre holding the BBVA We Think Big award","details":["My favourite espresso machine.","A snapshot from my PhD presentation at the V Jornadas Científicas de DATAI, University of Navarra.","I received BBVA’s We Think Big award in 2024 for my work on patents and my academic contributions."],"path":"/#espresso","links":[],"fields":[]},{"id":"section:contact","kind":"section","title":"Contact & profiles","summary":"Contact me on LinkedIn","details":["GitHub · Hugging Face · Scholar · LinkedIn · X · ORCID · University of Navarra"],"path":"/#profiles","links":[{"label":"GitHub","url":"https://github.com/detasar"},{"label":"Hugging Face","url":"https://huggingface.co/emre"},{"label":"Scholar","url":"https://scholar.google.com/citations?user=IZZM_JAAAAAJ"},{"label":"LinkedIn","url":"https://www.linkedin.com/in/detasar/"},{"label":"X","url":"https://x.com/mehanixs"},{"label":"ORCID","url":"https://orcid.org/0000-0002-7788-0478"},{"label":"University of Navarra","url":"https://www.unav.edu/web/instituto-de-ciencia-de-los-datos-e-inteligencia-artificial/equipo#:~:text=Enhancing%20Fairness%20in%20AI%20Systems"}],"fields":[]},{"id":"section:readers","kind":"section","title":"Index, graph and profile","summary":"Reading index · Knowledge graph · Profile","details":["Search research, Turkish patent and utility model applications, models and code repositories in one place.","Choose a research area to explore how the work connects and follow the links to its original sources.","I’m an engineer and applied scientist on BBVA’s AI Innovation team, part of AI Transformation. I’m doing a PhD at the University of Navarra on fairness in AI."],"path":"/guide/","links":[],"fields":[]},{"id":"topic:uncertainty","kind":"topic","title":"Uncertainty","summary":"Can intelligence know its limits?","details":["Conformal prediction, calibrated uncertainty, and the boundary between a prediction and a decision."],"path":"/graph/#topic-uncertainty","links":[],"fields":["confidence"]},{"id":"topic:fairness","kind":"topic","title":"Fairness","summary":"Who carries the uncertainty?","details":["How uncertainty estimates and referral to human review affect different people in decision systems."],"path":"/graph/#topic-fairness","links":[],"fields":["confidence"]},{"id":"topic:agents","kind":"topic","title":"Agent systems","summary":"How can you check what an agent actually did?","details":["Tools that connect automated actions with evidence, understandable progress, and human oversight."],"path":"/graph/#topic-agents","links":[],"fields":["agents"]},{"id":"topic:quantum","kind":"topic","title":"Quantum","summary":"What can a measurement tell us?","details":["Statistical calibration, anomaly detection, and adversarial evaluation around quantum measurements."],"path":"/graph/#topic-quantum","links":[],"fields":["quantum"]},{"id":"topic:invention","kind":"topic","title":"Invention","summary":"How small would a rescue robot need to be?","details":["Patent applications on rescue robotics, conversational interfaces, personal knowledge graphs and quantum methods."],"path":"/graph/#topic-invention","links":[],"fields":["invention"]},{"id":"topic:language","kind":"topic","title":"Language","summary":"How do machines understand us?","details":["Turkish speech recognition, language-model adaptation, and contributions to open multilingual research."],"path":"/graph/#topic-language","links":[],"fields":["language"]},{"id":"affiliation:bbva","kind":"affiliation","title":"BBVA","summary":"Applied Engineer / Scientist","details":["BBVA · AI Transformation · AI Innovation","I’m an engineer and applied scientist on BBVA’s AI Innovation team, part of AI Transformation. I’m doing a PhD at the University of Navarra on fairness in AI."],"path":"/profile/","links":[{"label":"LinkedIn","url":"https://www.linkedin.com/in/detasar/"}],"fields":[]},{"id":"affiliation:navarra","kind":"affiliation","title":"University of Navarra","summary":"Doctoral researcher in AI fairness","details":["University of Navarra"],"path":"/profile/","links":[{"label":"University of Navarra","url":"https://www.unav.edu/web/instituto-de-ciencia-de-los-datos-e-inteligencia-artificial/equipo#:~:text=Enhancing%20Fairness%20in%20AI%20Systems"}],"fields":[]},{"id":"photo:espresso","kind":"photo","title":"Emre beside an Elektra espresso machine","summary":"My favourite espresso machine.","details":["Emre beside an Elektra espresso machine"],"path":"/#espresso","links":[],"image":"/assets/emre-espresso.jpg","fields":[]},{"id":"photo:scholar","kind":"photo","title":"Emre presenting his PhD research at Navarra","summary":"A snapshot from my PhD presentation at the V Jornadas Científicas de DATAI, University of Navarra.","details":["Emre presenting his PhD research at Navarra"],"path":"/#scholar","links":[{"label":"University of Navarra","url":"https://www.unav.edu/documents/d/instituto-de-ciencia-de-los-datos-e-inteligencia-artificial/programa_jornadas_cientificas_datai_mayo_2026"}],"image":"/assets/emre-scholar.webp","fields":[]},{"id":"photo:we-think-big","kind":"photo","title":"We Think Big","summary":"I received BBVA’s We Think Big award in 2024 for my work on patents and my academic contributions.","details":["BBVA · 2024"],"path":"/#we-think-big","links":[],"image":"/assets/emre-we-think-big.jpg","fields":[],"year":"2024"},{"id":"note:ontologies-and-data-structures","kind":"note","title":"Ontologies and data structures","summary":"What changes in a graph when a train runs late?","details":["A station is a place. A departure is an event. If I give each its own node type and each time a source, can I trace exactly which connections a single delay affects?"],"path":"/#idea-notebook-title","links":[{"label":"RDF 1.1 · W3C","url":"https://www.w3.org/TR/rdf11-concepts/"}],"fields":["meaning"]},{"id":"note:category-theory","kind":"note","title":"Category theory","summary":"What survives when I compose two transformations?","details":["One process turns a timetable into a graph. Another finds a route through it. When I compose them, how do I preserve which station each time belongs to?"],"path":"/#idea-notebook-title","links":[{"label":"Seven Sketches in Compositionality","url":"https://arxiv.org/abs/1803.05316"}],"fields":["meaning"]},{"id":"note:quantum-measurement","kind":"note","title":"Quantum measurement","summary":"What if I ask the same qubit a different question?","details":["I can prepare a qubit so it always gives the same answer when measured along one axis. Along a perpendicular axis, either of two results is equally likely. What fascinates me is how changing the measurement changes the answer while the preparation stays the same."],"path":"/#idea-notebook-title","links":[{"label":"Quantum Information · IBM","url":"https://quantum.cloud.ibm.com/learning/en/courses/basics-of-quantum-information/single-systems/quantum-information"}],"fields":["quantum"]},{"id":"project:meaning-native-computing","kind":"project","title":"Meaning-Native Computing","summary":"Can a computer predict which instructions will run next if it knows the rules of the program?","details":["Meaning-Native Computing is my independent research project. I want to compare two predictors given the same program and comparable computing resources. One also gets an explicit description of the rules the program must follow. Does that help it predict the next steps? Before a predicted change can take effect, an independent check must confirm that it follows those rules."],"path":"/#meaning-native-computing","links":[],"meta":"Davut Emre Taşar","fields":["meaning","confidence"]},{"id":"work:coverage-deferral","kind":"work","title":"The coverage–deferral trade-off","summary":"A study of how conformal prediction changes both uncertainty coverage and who is referred for human review.","details":["Whose case gets a second look?","A prediction set can contain both possible answers. In this study, those cases go to a human reviewer. I compared a single calibration threshold with thresholds fitted separately for each group. The choice of threshold changes how often the right answer is in the set and how often a person has to wait for review.","On the Adult income dataset, groupwise calibration narrowed the gap in coverage from 5.0 to 1.2 percentage points. The gap in referral to a human grew from 5.1 to 12.9 points. Table 4 reports these results over 100 random seeds, with 90% target coverage and simulated review."],"path":"/work/coverage-deferral/","links":[{"label":"Original source","url":"https://www.preprints.org/manuscript/202512.2631"},{"label":"DOI","url":"https://doi.org/10.20944/preprints202512.2631.v1"}],"meta":"2025 · Preprint · Preprints.org · Sole author","identifier":"10.20944/preprints202512.2631.v1","fields":["confidence","learning"],"year":"2025"},{"id":"work:agent-chronicle","kind":"work","title":"Agent Chronicle","summary":"A local control surface for coding agents that connects progress, decisions, changes, and verification to recorded evidence.","details":["A coding agent says it is finished. What changed?","Agent Chronicle turns a coding session into a timeline you can inspect. Follow a Codex or Claude task from its original goal through decisions, file changes and checks, then open the recorded activity behind a claim.","The public repository includes a local demo, setup guides and an Apache 2.0 licence for the project’s own code. Its interface is available in English, Turkish and Spanish."],"path":"/work/agent-chronicle/","links":[{"label":"Original source","url":"https://github.com/detasar/agent-chronicle-community"},{"label":"Documentation","url":"https://github.com/detasar/agent-chronicle-community/tree/main/docs"}],"meta":"2026 · Public source project · Project maintainer","identifier":"detasar/agent-chronicle-community","fields":["agents"],"year":"2026"},{"id":"work:quantum-conformal","kind":"work","title":"Quantum-enhanced conformal methods","summary":"An investigation of conformal uncertainty estimation for multi-output distributions arising from quantum measurements.","details":["How much can a finite number of measurements tell us?","A quantum experiment produces counts for several possible outcomes. I studied how to estimate their probabilities together and how wide a set of plausible answers should be. The method calibrates predictions against earlier examples and checks the result on held-out data.","The 2025 journal article evaluates both coverage and set size on simulated data and measurements in several bases. The accompanying notebooks show how to rerun the experiments."],"path":"/work/quantum-conformal/","links":[{"label":"Original source","url":"https://journals.gen.tr/index.php/jqtair/article/view/2702"},{"label":"Explore code","url":"https://github.com/detasar/QECMMOU"},{"label":"DOI","url":"https://doi.org/10.70447/ktve.2702"}],"meta":"2025 · Published journal article · Sole author","identifier":"10.70447/ktve.2702","fields":["quantum","confidence"],"year":"2025"},{"id":"work:tara","kind":"work","title":"TARA · Test by Adaptive Ranks","summary":"A framework combining conformal prediction and sequential testing to study anomalies in quantum correlations.","details":["When does a stream of measurements start to look unusual?","Ceren Öcal Taşar and I studied how to flag changes in quantum correlations as observations arrive. One version compares a batch with a reference distribution. A second accumulates evidence over successive measurements.","The preprint reports experiments on IBM Torino and IonQ Forte Enterprise. It also examines how using the same distribution for calibration and evaluation can change the apparent performance."],"path":"/work/tara/","links":[{"label":"Original source","url":"https://arxiv.org/abs/2512.04016"},{"label":"PDF","url":"https://arxiv.org/pdf/2512.04016"}],"meta":"2025 · arXiv preprint · First author","identifier":"arXiv:2512.04016","fields":["quantum","confidence"],"year":"2025"},{"id":"work:chat-assistant","kind":"work","title":"Conversational systems that act","summary":"An invention proposal for an AI-assisted conversational system combining information retrieval with command-driven actions.","details":["When should a conversation turn into an action?","The proposed assistant can search for information when a user asks a question and carry out an operation when given a command. The application puts both kinds of interaction in a single conversation.","The Turkish filing is TR 2023/011870. The international application was published on 3 April 2025."],"path":"/work/chat-assistant/","links":[{"label":"Original source","url":"https://patents.google.com/patent/WO2025071524A1/en"},{"label":"Turkish publication","url":"https://patents.google.com/patent/TR2023011870A2/tr"}],"meta":"2025 · Published PCT application · Co-inventor","identifier":"WO2025071524A1","fields":["invention","agents","language"],"year":"2025"},{"id":"work:micro-robot-rescue","kind":"work","title":"Micro-robots for search and rescue","summary":"A proposed micro-robot system for locating victims and supporting rescue operations in confined environments.","details":["What could reach someone through a gap only a finger wide?","Our application proposes a finger-sized, jointed robot for narrow spaces in rubble. A camera and sensors would help locate a trapped person. A tether would carry a communications and support line through the same opening.","Türkiye Garanti Bankası filed TR 2023/011901 on 25 September 2023. The international publication WO2025071525A1 followed on 3 April 2025."],"path":"/work/micro-robot-rescue/","links":[{"label":"Original source","url":"https://patents.google.com/patent/WO2025071525A1/en"},{"label":"Turkish publication","url":"https://patents.google.com/patent/TR2023011901A2/tr"}],"meta":"2025 · Published PCT application · Co-inventor","identifier":"WO2025071525A1","image":"/assets/rescue-patent-figure.webp","fields":["invention","agents"],"year":"2025"},{"id":"work:ai-pos","kind":"work","title":"AI-assisted point of sale","summary":"A proposed POS system connecting personalized recommendations, fraud detection, and natural-language device operation.","details":["Could a shopkeeper operate a payment terminal by speaking?","This application describes a merchant operating a POS terminal remotely with spoken or written instructions. It also proposes generating QR-coded output and analysing transactions for potential fraud.","The international application WO2024220062A1 was published on 24 October 2024. The related Turkish filing is TR 2023/004244."],"path":"/work/ai-pos/","links":[{"label":"Original source","url":"https://patents.google.com/patent/WO2024220062A1/en"},{"label":"Turkish publication","url":"https://patents.google.com/patent/TR2023004244A2/tr"}],"meta":"2024 · Published PCT application · Co-inventor","identifier":"WO2024220062A1","fields":["invention","language"],"year":"2024"},{"id":"work:virtual-input","kind":"work","title":"Gesture-based virtual interaction","summary":"An invention proposal using hand-motion tracking for interaction with virtual and augmented reality environments.","details":["Where is the keyboard when the screen is virtual?","The proposal uses camera observations of hand movement as input in virtual and augmented reality. The system interprets those movements as keyboard or mouse actions.","The Turkish record TR 2023/011601 and the PCT publication WO2025063923A1 describe the same invention. The PCT application was published on 27 March 2025."],"path":"/work/virtual-input/","links":[{"label":"Original source","url":"https://patents.google.com/patent/WO2025063923A1/en"},{"label":"Turkish publication","url":"https://patents.google.com/patent/TR2023011601A2/tr"}],"meta":"2025 · Published PCT application · Co-inventor","identifier":"WO2025063923A1","fields":["invention"],"year":"2025"},{"id":"work:personalized-suggestions","kind":"work","title":"Personalized workplace suggestions","summary":"A proposed system generating personalized workplace recommendations from daily mood and health signals.","details":["How could a workplace assistant tailor a suggestion?","The application describes a system that generates suggestions using signals about an employee’s day, mood and health. The document sets out a proposed design for workplace recommendations.","The related Turkish filing is TR 2023/011830. The international publication WO2025063924A1 lists Türkiye Garanti Bankası as applicant."],"path":"/work/personalized-suggestions/","links":[{"label":"Original source","url":"https://patents.google.com/patent/WO2025063924A1/en"},{"label":"Turkish publication","url":"https://patents.google.com/patent/TR2023011830A2/tr"}],"meta":"2025 · Published PCT application · Co-inventor","identifier":"WO2025063924A1","fields":["invention"],"year":"2025"},{"id":"work:adversarial-quantum","kind":"work","title":"Adversarial limits of quantum certification","summary":"An exploration of adaptive adversaries and evaluation limits in statistical quantum certification.","details":["Does a good test result hold up under different conditions?","I studied how the choice of calibration and evaluation data affects reported performance in tests of quantum correlations. The experiment asks when a statistical score is a reliable piece of evidence.","The December 2025 preprint describes the simulation, the comparison with IBM Quantum measurements and the assumptions behind the evaluation."],"path":"/work/adversarial-quantum/","links":[{"label":"Original source","url":"https://arxiv.org/abs/2512.04391"},{"label":"PDF","url":"https://arxiv.org/pdf/2512.04391"}],"meta":"2025 · arXiv preprint · Sole author","identifier":"arXiv:2512.04391","fields":["quantum","confidence"],"year":"2025"},{"id":"work:turkish-asr","kind":"work","title":"Turkish speech recognition","summary":"An evaluation of self-supervised speech representations and transfer learning for Turkish speech recognition.","details":["What carries over when a speech model learns Turkish?","With Kutan Koruyan and Cihan Çılgın, I adapted a Wav2Vec 2.0 model pretrained on speech in 53 languages. We used recordings and transcripts from Mozilla Common Voice Turkish to teach the model to turn Turkish speech into text.","The paper reports a 23% word error rate on the Common Voice Turkish test split. This measure counts substitutions, deletions and insertions relative to the length of the reference transcript."],"path":"/work/turkish-asr/","links":[{"label":"Original source","url":"https://dergipark.org.tr/en/pub/acin/article/1338604"},{"label":"DOI","url":"https://doi.org/10.26650/acin.1338604"}],"meta":"2024 · Published journal article · First author","identifier":"10.26650/acin.1338604","fields":["language"],"year":"2024"},{"id":"work:qwen-grpo","kind":"work","title":"Qwen-0.5B-GRPO","summary":"A GRPO fine-tune of Qwen2.5-0.5B-Instruct for structured mathematical reasoning using GSM8K.","details":["Can a small model learn to lay out a maths answer?","I adapted Qwen2.5-0.5B-Instruct on GSM8K word problems. The fine-tuning rewards a correct numerical answer and a response that follows the requested structure. These properties are scored separately.","The model card documents a single training epoch, the upstream Qwen model and example inference code. The public weights let other researchers try the adaptation themselves."],"path":"/work/qwen-grpo/","links":[{"label":"Original source","url":"https://huggingface.co/emre/Qwen-0.5B-GRPO"},{"label":"Base model","url":"https://huggingface.co/Qwen/Qwen2.5-0.5B-Instruct"}],"meta":"2025 · Public model adaptation · Model developer","identifier":"emre/Qwen-0.5B-GRPO","fields":["language"],"year":"2025"},{"id":"work:turkish-reasoning-benchmark","kind":"work","title":"TARA · Reasoning in Turkish","summary":"A set of Turkish-language tasks for examining how language models reason across domains.","details":["How do you ask a language model to reason in Turkish?","Ceren Öcal Taşar and I assembled tasks across ten categories, including logic, planning, Python and SQL. A reader can inspect the prompts and expected responses before looking at a model’s score.","Version 1 contains 100 synthetic examples. The dataset card describes their generation and notes that a single generator and a small sample limit what the scores say about everyday Turkish."],"path":"/work/turkish-reasoning-benchmark/","links":[{"label":"Original source","url":"https://huggingface.co/datasets/emre/TARA_Turkish_LLM_Benchmark"},{"label":"DOI","url":"https://doi.org/10.57967/hf/5193"}],"meta":"2025 · Public research dataset · Version 1 · Co-creator","identifier":"10.57967/hf/5193","fields":["language"],"year":"2025"},{"id":"work:multilingual-chatbots","kind":"work","title":"One language model for several chatbots","summary":"We tested whether separate chatbots could share one multilingual BERT model while keeping their own sets of possible answers.","details":["Do thirteen chatbot tasks need thirteen models?","A visitor asking how to get a virtual card and a visitor asking about an unrelated subject need different answers. Our classifier shares the language model across chatbots and uses the active chatbot to narrow the possible responses. The evaluation included banking questions covering 77 intents.","Across the study we tested 13 classification tasks and 333 classes in Turkish, English and German. In the five-model comparison, the separate models used 10 GB of memory in total. The shared model used 2 GB, with a small loss of accuracy."],"path":"/work/multilingual-chatbots/","links":[{"label":"Original source","url":"https://dergipark.org.tr/en/pub/jaida/article/1060221"},{"label":"Full text","url":"https://dergipark.org.tr/tr/download/article-file/2201572"}],"meta":"2021 · Peer-reviewed journal article · First and corresponding author","identifier":"IZ:JA77UY82ER","fields":["language"],"year":"2021"},{"id":"work:ad-text-classification","kind":"work","title":"Finding mismatched labels in ad copy","summary":"We examined how mismatches between an advertisement and its industry label affect a BERT text classifier.","details":["What if the label on an ad is the mistake?","The classifier struggled in particular industries. We used a zero-shot model to check whether each ad matched its assigned sector, cleaned the training data, then trained BERT again.","The original corpus contained about 101,000 labelled ads from 44 industries. The article reports improved classification performance after this preprocessing step."],"path":"/work/ad-text-classification/","links":[{"label":"Original source","url":"https://dergipark.org.tr/tr/pub/tbbmd/article/1004781"},{"label":"DOI","url":"https://doi.org/10.54525/tbbmd.1004781"}],"meta":"2021 · Peer-reviewed journal article · Co-author","identifier":"10.54525/tbbmd.1004781","fields":["language"],"year":"2021"},{"id":"work:weight-editing","kind":"work","title":"Editing neural networks with language models","summary":"I asked language models to propose small weight changes for two different tasks. One target was a damaged classifier. The other was an intact language model.","details":["Can one model edit the weights of another?","In the UCI Adult income classifier I permanently disabled some weights. Sol and Astra could propose edits to the surviving weights, then a joint search could use their development histories. In a separate experiment I tested edits on intact SmolLM2-1.7B using ARC-Easy science questions and PIQA physical-reasoning questions. In both experiments, keeping the original weights was an eligible option.","On 16,281 held-out Adult examples, Sol’s edit raised balanced accuracy from 78.59% to 79.02%. Overall accuracy fell from 82.86% to 82.63%, and probability estimates worsened. On SmolLM2 every completed search selected the original weights. These are exploratory results from one run.","Damaged classifier · UCI Adult · Sol","Balanced accuracy · Baseline 78.59% · Selected 79.02%","Overall accuracy · Baseline 82.86% · Selected 82.63%","Intact language model · SmolLM2-1.7B · Original weights selected"],"path":"/work/weight-editing/","links":[{"label":"Original source","url":"https://github.com/detasar/llm-weight-editing-study"},{"label":"Research report","url":"https://github.com/detasar/llm-weight-editing-study/blob/main/publication/2026-11-04/RESEARCH_REPORT.pdf"},{"label":"Interactive presentation","url":"https://detasar.github.io/llm-weight-editing-study/publication/2026-11-04/START_PRESENTATION.html"},{"label":"Measurements","url":"https://github.com/detasar/llm-weight-editing-study/blob/main/data/manifest.json"}],"meta":"2026 · Public experiment, report and code · Sole author","identifier":"detasar/llm-weight-editing-study","fields":["learning","agents","language"],"year":"2026"},{"id":"work:bloom","kind":"work","title":"BLOOM","summary":"A contribution to the BigScience collaboration on open multilingual language-model research.","details":["How was a multilingual model built in the open?","BigScience brought together a large international research collaboration to document the training and release of BLOOM, a multilingual language model. I am one of the co-authors of the project paper.","The BigScience paper gives the full author list and describes the architecture, data and training process."],"path":"/work/bloom/","links":[{"label":"Original source","url":"https://arxiv.org/abs/2211.05100"}],"meta":"2022 · Collaborative research paper · Co-author","identifier":"arXiv:2211.05100","fields":["language"],"year":"2022"},{"id":"patent:2026-007909","kind":"patent","title":"Noise reduction system based on the user's voice","summary":"The system isolates a speaker’s voice from background sounds and transcribes it. It can edit the wording to suit the speaker’s preferences or the setting, then regenerate the speech with the speaker’s vocal characteristics.","details":["Filed · May 18, 2026","Applicant · TÜRKİYE GARANTİ BANKASI ANONİM ŞİRKETİ","Published · Jul 21, 2026","Portfolio-wide legal status is not provided by the dated extract. Individual grant decisions cite the national register."],"path":"/patents/2026-007909/","links":[{"label":"Original source","url":"https://www.turkpatent.gov.tr/arastirma-yap?form=patent"}],"meta":"TR 2026/007909 · Patent application","fields":["invention","language"],"year":"2026"},{"id":"patent:2026-007907","kind":"patent","title":"Logical inference-based artificial intelligence system","summary":"The application describes turning transaction histories, user records and network connections into a logical knowledge model. The system can then draw inferences from those records in fields such as finance, health and logistics.","details":["Filed · May 18, 2026","Applicant · TÜRKİYE GARANTİ BANKASI ANONİM ŞİRKETİ","Published · Jun 22, 2026","Portfolio-wide legal status is not provided by the dated extract. Individual grant decisions cite the national register."],"path":"/patents/2026-007907/","links":[{"label":"Original source","url":"https://www.turkpatent.gov.tr/arastirma-yap?form=patent"}],"meta":"TR 2026/007907 · Patent application","fields":["meaning","invention"],"year":"2026"},{"id":"patent:2026-006294","kind":"patent","title":"A personalized waiting experience system","summary":"The proposed system assesses a person’s emotional state while they wait and selects content for them based on that assessment.","details":["Filed · Apr 24, 2026","Applicant · TÜRKİYE GARANTİ BANKASI ANONİM ŞİRKETİ","Published · May 21, 2026","Portfolio-wide legal status is not provided by the dated extract. Individual grant decisions cite the national register."],"path":"/patents/2026-006294/","links":[{"label":"Original source","url":"https://www.turkpatent.gov.tr/arastirma-yap?form=patent"}],"meta":"TR 2026/006294 · Patent application","fields":["invention"],"year":"2026"},{"id":"patent:2026-006289","kind":"patent","title":"A holistic retirement planning and life assistance system supported by a personalized dynamic ontology and knowledge graph","summary":"The application proposes a personal ontology and a knowledge graph that can evolve over time. It brings those representations together to support retirement planning and offer proactive assistance in everyday life.","details":["Filed · Apr 24, 2026","Applicant · TÜRKİYE GARANTİ BANKASI ANONİM ŞİRKETİ","Published · Jun 22, 2026","Portfolio-wide legal status is not provided by the dated extract. Individual grant decisions cite the national register."],"path":"/patents/2026-006289/","links":[{"label":"Original source","url":"https://www.turkpatent.gov.tr/arastirma-yap?form=patent"}],"meta":"TR 2026/006289 · Patent application","fields":["meaning","invention"],"year":"2026"},{"id":"patent:2026-004444","kind":"patent","title":"An AI-supported and privacy-focused environment sensing system","summary":"The application addresses spaces where cameras are restricted or unsuitable. It proposes detecting whether people are present or active while preserving privacy, so that energy use can be adjusted to the situation.","details":["Filed · Mar 26, 2026","Applicant · TÜRKİYE GARANTİ BANKASI ANONİM ŞİRKETİ","Published · May 21, 2026","Portfolio-wide legal status is not provided by the dated extract. Individual grant decisions cite the national register."],"path":"/patents/2026-004444/","links":[{"label":"Original source","url":"https://www.turkpatent.gov.tr/arastirma-yap?form=patent"}],"meta":"TR 2026/004444 · Patent application","fields":["invention"],"year":"2026"},{"id":"patent:2026-004441","kind":"patent","title":"A noise-resistant dynamic financial customer segmentation system using quantum entanglement and conformal prediction","summary":"The application proposes dynamic financial customer segments that can tolerate noisy data. It brings together quantum entanglement to represent complex patterns, variational quantum clustering to form groups, and conformal prediction to provide statistical assurance.","details":["Filed · Mar 26, 2026","Applicant · TÜRKİYE GARANTİ BANKASI ANONİM ŞİRKETİ","Published · Apr 21, 2026","Portfolio-wide legal status is not provided by the dated extract. Individual grant decisions cite the national register."],"path":"/patents/2026-004441/","links":[{"label":"Original source","url":"https://www.turkpatent.gov.tr/arastirma-yap?form=patent"}],"meta":"TR 2026/004441 · Patent application","fields":["invention","quantum","confidence"],"year":"2026"},{"id":"patent:2026-002909","kind":"patent","title":"A prediction system","summary":"The system is intended to give a numerical measure of confidence in an AI prediction. The abstract gives examples including a borrower defaulting, a transaction producing a profit or loss, and an operation finishing late.","details":["Filed · Feb 26, 2026","Applicant · TÜRKİYE GARANTİ BANKASI ANONİM ŞİRKETİ","Published · Mar 23, 2026","Portfolio-wide legal status is not provided by the dated extract. Individual grant decisions cite the national register."],"path":"/patents/2026-002909/","links":[{"label":"Original source","url":"https://www.turkpatent.gov.tr/arastirma-yap?form=patent"}],"meta":"TR 2026/002909 · Patent application","fields":["invention"],"year":"2026"},{"id":"patent:2026-001992","kind":"patent","title":"A system providing customer identification, time measurement and experience optimization in banks using computer imagery and artificial intelligence","summary":"The proposed system uses existing computer-vision and AI algorithms to identify customers in a bank branch, measure visit durations and estimate waiting times. It also describes monitoring departures and adjusting the queue.","details":["Filed · Feb 10, 2026","Applicant · TÜRKİYE GARANTİ BANKASI ANONİM ŞİRKETİ","Published · Mar 23, 2026","Portfolio-wide legal status is not provided by the dated extract. Individual grant decisions cite the national register."],"path":"/patents/2026-001992/","links":[{"label":"Original source","url":"https://www.turkpatent.gov.tr/arastirma-yap?form=patent"}],"meta":"TR 2026/001992 · Patent application","fields":["invention"],"year":"2026"},{"id":"patent:2025-019989","kind":"patent","title":"A gamification system","summary":"The system analyses banking activity and proposes generating financial content and game scenarios tailored to the user. The banking application can collect feedback on those scenarios.","details":["Filed · Dec 12, 2025","Applicant · TÜRKİYE GARANTİ BANKASI ANONİM ŞİRKETİ","Published · Jan 21, 2026","TR2025019989A2","Portfolio-wide legal status is not provided by the dated extract. Individual grant decisions cite the national register."],"path":"/patents/2025-019989/","links":[{"label":"TR2025019989A2","url":"https://patents.google.com/patent/TR2025019989A2/en"},{"label":"Official TÜRKPATENT register","url":"https://www.turkpatent.gov.tr/arastirma-yap?form=patent"}],"meta":"TR 2025/019989 · Patent application","fields":["invention"],"year":"2025"},{"id":"patent:2025-011133","kind":"patent","title":"A system enabling cognitive enrichment of captured image content with federated learning support","summary":"The filing describes personalising audiovisual memories captured on a user’s device, with federated learning as part of the proposed design.","details":["Filed · Aug 8, 2025","Applicant · TÜRKİYE GARANTİ BANKASI ANONİM ŞİRKETİ","Published · Sep 22, 2025","TR2025011133A2","Portfolio-wide legal status is not provided by the dated extract. Individual grant decisions cite the national register."],"path":"/patents/2025-011133/","links":[{"label":"TR2025011133A2","url":"https://patents.google.com/patent/TR2025011133A2/en"},{"label":"Official TÜRKPATENT register","url":"https://www.turkpatent.gov.tr/arastirma-yap?form=patent"}],"meta":"TR 2025/011133 · Patent application","fields":["learning","invention"],"year":"2025"},{"id":"patent:2025-008786","kind":"patent","title":"A probabilistic and dynamically extensible graph-based variant generation system for multimodal content","summary":"The system extracts features from sound, video or images, then builds a graph of possible variations. Branches can represent alternative pasts, futures or parallel versions of the original content.","details":["Filed · Jun 30, 2025","Applicant · TÜRKİYE GARANTİ BANKASI ANONİM ŞİRKETİ","Published · Aug 21, 2025","TR2025008786A2","Portfolio-wide legal status is not provided by the dated extract. Individual grant decisions cite the national register."],"path":"/patents/2025-008786/","links":[{"label":"TR2025008786A2","url":"https://patents.google.com/patent/TR2025008786A2/en"},{"label":"Official TÜRKPATENT register","url":"https://www.turkpatent.gov.tr/arastirma-yap?form=patent"}],"meta":"TR 2025/008786 · Patent application","fields":["meaning","invention"],"year":"2025"},{"id":"patent:2025-008273","kind":"patent","title":"A personal transaction booth system","summary":"A modular service booth gives a person a private place to handle sensitive interactions with several service providers.","details":["Filed · Jun 20, 2025","Applicant · TÜRKİYE GARANTİ BANKASI ANONİM ŞİRKETİ","Published · Sep 22, 2025","TR2025008273A2","Portfolio-wide legal status is not provided by the dated extract. Individual grant decisions cite the national register."],"path":"/patents/2025-008273/","links":[{"label":"TR2025008273A2","url":"https://patents.google.com/patent/TR2025008273A2/en"},{"label":"Official TÜRKPATENT register","url":"https://www.turkpatent.gov.tr/arastirma-yap?form=patent"}],"meta":"TR 2025/008273 · Patent application","fields":["invention"],"year":"2025"},{"id":"patent:2025-008055","kind":"patent","title":"A driving safety system","summary":"The proposed system uses sensors and AI to detect and assess a vehicle collision as it occurs. It can notify the relevant services and suggest transport or accommodation to the people involved.","details":["Filed · Jun 18, 2025","Applicant · TÜRKİYE GARANTİ BANKASI ANONİM ŞİRKETİ","Published · Aug 21, 2025","Portfolio-wide legal status is not provided by the dated extract. Individual grant decisions cite the national register."],"path":"/patents/2025-008055/","links":[{"label":"Original source","url":"https://www.turkpatent.gov.tr/arastirma-yap?form=patent"}],"meta":"TR 2025/008055 · Patent application","fields":["invention"],"year":"2025"},{"id":"patent:2025-008032","kind":"patent","title":"A collective decision-making system using human–AI hybrid swarm intelligence","summary":"The application brings human and AI participants into a collective decision process. It proposes several levels of reliability assessment within a system designed to adapt and improve over time.","details":["Filed · Jun 17, 2025","Applicant · TÜRKİYE GARANTİ BANKASI ANONİM ŞİRKETİ","Published · Jul 21, 2025","Portfolio-wide legal status is not provided by the dated extract. Individual grant decisions cite the national register."],"path":"/patents/2025-008032/","links":[{"label":"Original source","url":"https://www.turkpatent.gov.tr/arastirma-yap?form=patent"}],"meta":"TR 2025/008032 · Patent application","fields":["invention"],"year":"2025"},{"id":"patent:2025-008028","kind":"patent","title":"A coordination system in a multi-device ecosystem","summary":"The proposed system uses AI to coordinate devices and applications within an ecosystem. Its stated aim is a user experience that can evolve as the person interacts with that ecosystem.","details":["Filed · Jun 17, 2025","Applicant · TÜRKİYE GARANTİ BANKASI ANONİM ŞİRKETİ","Published · Jul 21, 2025","Portfolio-wide legal status is not provided by the dated extract. Individual grant decisions cite the national register."],"path":"/patents/2025-008028/","links":[{"label":"Original source","url":"https://www.turkpatent.gov.tr/arastirma-yap?form=patent"}],"meta":"TR 2025/008028 · Patent application","fields":["invention"],"year":"2025"},{"id":"patent:2025-007997","kind":"patent","title":"A holistic real estate valuation analysis and strategic planning system","summary":"This application combines data from multiple sources for real-estate valuation. It proposes using an evolving ontology, network analysis and explainable generative AI to connect valuation with analysis and strategic planning.","details":["Filed · Jun 17, 2025","Applicant · TÜRKİYE GARANTİ BANKASI ANONİM ŞİRKETİ","Published · Jul 21, 2025","Portfolio-wide legal status is not provided by the dated extract. Individual grant decisions cite the national register."],"path":"/patents/2025-007997/","links":[{"label":"Original source","url":"https://www.turkpatent.gov.tr/arastirma-yap?form=patent"}],"meta":"TR 2025/007997 · Patent application","fields":["invention"],"year":"2025"},{"id":"patent:2025-007995","kind":"patent","title":"A development environment system with an agent architecture and related data format","summary":"The application proposes an agent architecture and an accompanying data format for building AI assistants. The agents are designed to move between platforms and adapt over time while retaining a personalized experience.","details":["Filed · Jun 17, 2025","Applicant · TÜRKİYE GARANTİ BANKASI ANONİM ŞİRKETİ","Published · Aug 21, 2025","Portfolio-wide legal status is not provided by the dated extract. Individual grant decisions cite the national register."],"path":"/patents/2025-007995/","links":[{"label":"Original source","url":"https://www.turkpatent.gov.tr/arastirma-yap?form=patent"}],"meta":"TR 2025/007995 · Patent application","fields":["invention","agents"],"year":"2025"},{"id":"patent:2024-003815","kind":"patent","title":"A system improving customer communication in the financial sector","summary":"The proposal connects mobile banking, augmented reality and virtual branches so customers can discuss investments or other financial topics with each other and interact with AI models.","details":["Filed · Mar 27, 2024","Applicant · TÜRKİYE GARANTİ BANKASI ANONİM ŞİRKETİ","Published · Apr 22, 2024","TR2024003815A2","Portfolio-wide legal status is not provided by the dated extract. Individual grant decisions cite the national register."],"path":"/patents/2024-003815/","links":[{"label":"TR2024003815A2","url":"https://patents.google.com/patent/TR2024003815A2/en"},{"label":"Official TÜRKPATENT register","url":"https://www.turkpatent.gov.tr/arastirma-yap?form=patent"}],"meta":"TR 2024/003815 · Patent application","fields":["invention"],"year":"2024"},{"id":"patent:2024-003813","kind":"patent","title":"A system enabling the creation of investment channels","summary":"The application proposes investment channels curated by experts or other public figures. It describes AI avatars based on particular profiles delivering financial guidance to the channel’s users.","details":["Filed · Mar 27, 2024","Applicant · TÜRKİYE GARANTİ BANKASI ANONİM ŞİRKETİ","Published · Apr 22, 2024","Portfolio-wide legal status is not provided by the dated extract. Individual grant decisions cite the national register."],"path":"/patents/2024-003813/","links":[{"label":"Original source","url":"https://www.turkpatent.gov.tr/arastirma-yap?form=patent"}],"meta":"TR 2024/003813 · Patent application","fields":["invention"],"year":"2024"},{"id":"patent:2024-003592","kind":"patent","title":"A system enabling people to help one another with certain banking transactions through mutual consent","summary":"A customer who wants help with a banking transaction can complete it with someone they trust. The proposal uses mutual approval to let the two people take part together.","details":["Filed · Mar 25, 2024","Applicant · TÜRKİYE GARANTİ BANKASI ANONİM ŞİRKETİ","Published · May 21, 2024","TR2024003592A2","Portfolio-wide legal status is not provided by the dated extract. Individual grant decisions cite the national register."],"path":"/patents/2024-003592/","links":[{"label":"TR2024003592A2","url":"https://patents.google.com/patent/TR2024003592A2/en"},{"label":"Official TÜRKPATENT register","url":"https://www.turkpatent.gov.tr/arastirma-yap?form=patent"}],"meta":"TR 2024/003592 · Patent application","fields":["invention"],"year":"2024"},{"id":"patent:2023-011901","kind":"patent","title":"A system providing search and rescue operations through a micro-robot","summary":"Micro-robots would enter confined spaces in a collapsed building or mine, gather information about the surroundings and help locate trapped people. The system also proposes a line for delivering support.","details":["Filed · Sep 25, 2023","Applicant · TÜRKİYE GARANTİ BANKASI ANONİM ŞİRKETİ","Published · Dec 21, 2023","TR2023011901A2","Final Turkish patent-grant decision announced · Feb 23, 2026","Portfolio-wide legal status is not provided by the dated extract. Individual grant decisions cite the national register."],"path":"/patents/2023-011901/","links":[{"label":"TR2023011901A2","url":"https://patents.google.com/patent/TR2023011901A2/en"},{"label":"Official TÜRKPATENT register","url":"https://www.turkpatent.gov.tr/arastirma-yap?form=patent"}],"meta":"TR 2023/011901 · Patent application","image":"/assets/rescue-patent-figure.webp","fields":["invention"],"year":"2023"},{"id":"patent:2023-011870","kind":"patent","title":"A chat assistant system","summary":"The proposed assistant can look up information online when a user asks a question and perform an operation when the user gives a command. Both interactions take place in the same conversation.","details":["Filed · Sep 25, 2023","Applicant · TÜRKİYE GARANTİ BANKASI ANONİM ŞİRKETİ","Published · Dec 21, 2023","TR2023011870A2","Portfolio-wide legal status is not provided by the dated extract. Individual grant decisions cite the national register."],"path":"/patents/2023-011870/","links":[{"label":"TR2023011870A2","url":"https://patents.google.com/patent/TR2023011870A2/en"},{"label":"Official TÜRKPATENT register","url":"https://www.turkpatent.gov.tr/arastirma-yap?form=patent"}],"meta":"TR 2023/011870 · Patent application","fields":["invention","language"],"year":"2023"},{"id":"patent:2023-011830","kind":"patent","title":"A system increasing employee productivity through personalized suggestions","summary":"The proposed workplace assistant uses daily information about an employee’s mood and health to generate individual suggestions for the working day.","details":["Filed · Sep 22, 2023","Applicant · TÜRKİYE GARANTİ BANKASI ANONİM ŞİRKETİ","Published · Aug 21, 2024","TR2023011830A2","Portfolio-wide legal status is not provided by the dated extract. Individual grant decisions cite the national register."],"path":"/patents/2023-011830/","links":[{"label":"TR2023011830A2","url":"https://patents.google.com/patent/TR2023011830A2/en"},{"label":"Official TÜRKPATENT register","url":"https://www.turkpatent.gov.tr/arastirma-yap?form=patent"}],"meta":"TR 2023/011830 · Patent application","fields":["invention"],"year":"2023"},{"id":"patent:2023-011601","kind":"patent","title":"A system enabling use of a virtual keyboard and mouse","summary":"The system tracks hand movements and interprets them as keyboard or mouse input in virtual and augmented reality.","details":["Filed · Sep 18, 2023","Applicant · TÜRKİYE GARANTİ BANKASI ANONİM ŞİRKETİ","Published · Dec 21, 2023","TR2023011601A2","Portfolio-wide legal status is not provided by the dated extract. Individual grant decisions cite the national register."],"path":"/patents/2023-011601/","links":[{"label":"TR2023011601A2","url":"https://patents.google.com/patent/TR2023011601A2/en"},{"label":"Official TÜRKPATENT register","url":"https://www.turkpatent.gov.tr/arastirma-yap?form=patent"}],"meta":"TR 2023/011601 · Patent application","fields":["invention"],"year":"2023"},{"id":"patent:2023-011600","kind":"patent","title":"A system increasing security in ATM transactions","summary":"The system watches for a card, cash or receipt left at an ATM after a transaction. The banking app can notify a customer who has walked away.","details":["Filed · Sep 18, 2023","Applicant · TÜRKİYE GARANTİ BANKASI ANONİM ŞİRKETİ","Published · Jan 22, 2024","TR2023011600A2","Portfolio-wide legal status is not provided by the dated extract. Individual grant decisions cite the national register."],"path":"/patents/2023-011600/","links":[{"label":"TR2023011600A2","url":"https://patents.google.com/patent/TR2023011600A2/en"},{"label":"Official TÜRKPATENT register","url":"https://www.turkpatent.gov.tr/arastirma-yap?form=patent"}],"meta":"TR 2023/011600 · Utility model application","fields":["invention"],"year":"2023"},{"id":"patent:2023-011599","kind":"patent","title":"A system facilitating ATM use","summary":"A phone or tablet reads a code on an ATM and displays its controls in augmented reality. The physical machine handles functions such as accepting and dispensing cash and printing receipts.","details":["Filed · Sep 18, 2023","Applicant · TÜRKİYE GARANTİ BANKASI ANONİM ŞİRKETİ","Published · Jan 22, 2024","TR2023011599A2","Portfolio-wide legal status is not provided by the dated extract. Individual grant decisions cite the national register."],"path":"/patents/2023-011599/","links":[{"label":"TR2023011599A2","url":"https://patents.google.com/patent/TR2023011599A2/en"},{"label":"Official TÜRKPATENT register","url":"https://www.turkpatent.gov.tr/arastirma-yap?form=patent"}],"meta":"TR 2023/011599 · Patent application","fields":["invention"],"year":"2023"},{"id":"patent:2023-004244","kind":"patent","title":"An AI-supported point-of-sale device management system","summary":"The proposal lets a merchant control a payment terminal remotely using speech or written instructions. It also describes QR-coded output and analysis of transactions for signs of fraud.","details":["Filed · Apr 17, 2023","Applicant · TÜRKİYE GARANTİ BANKASI ANONİM ŞİRKETİ","Published · Sep 21, 2023","TR2023004244A2","Portfolio-wide legal status is not provided by the dated extract. Individual grant decisions cite the national register."],"path":"/patents/2023-004244/","links":[{"label":"TR2023004244A2","url":"https://patents.google.com/patent/TR2023004244A2/en"},{"label":"Official TÜRKPATENT register","url":"https://www.turkpatent.gov.tr/arastirma-yap?form=patent"}],"meta":"TR 2023/004244 · Patent application","fields":["invention"],"year":"2023"},{"id":"patent:2023-004047","kind":"patent","title":"A secure shopping system","summary":"Parties to a credit or letter-of-credit transaction can create or join a contract. The system records its stages and supporting documents on a blockchain and uses AI to suggest matches between the parties.","details":["Filed · Apr 12, 2023","Applicant · TÜRKİYE GARANTİ BANKASI ANONİM ŞİRKETİ","Published · Sep 21, 2023","TR2023004047A2","Portfolio-wide legal status is not provided by the dated extract. Individual grant decisions cite the national register."],"path":"/patents/2023-004047/","links":[{"label":"TR2023004047A2","url":"https://patents.google.com/patent/TR2023004047A2/en"},{"label":"Official TÜRKPATENT register","url":"https://www.turkpatent.gov.tr/arastirma-yap?form=patent"}],"meta":"TR 2023/004047 · Patent application","fields":["invention"],"year":"2023"},{"id":"patent:2022-016679","kind":"patent","title":"A recommendation system","summary":"The proposal derives product suggestions from text, images, audio or movement supplied in a conversation or virtual environment. It keeps previous inputs and recommendations so the exchange can continue across interactions.","details":["Filed · Nov 3, 2022","Applicant · TÜRKİYE GARANTİ BANKASI ANONİM ŞİRKETİ","Published · Dec 21, 2022","WO2024096828A1","Portfolio-wide legal status is not provided by the dated extract. Individual grant decisions cite the national register."],"path":"/patents/2022-016679/","links":[{"label":"WO2024096828A1","url":"https://patents.google.com/patent/WO2024096828A1/en"},{"label":"Official TÜRKPATENT register","url":"https://www.turkpatent.gov.tr/arastirma-yap?form=patent"}],"meta":"TR 2022/016679 · Utility model application","fields":["invention"],"year":"2022"},{"id":"patent:2022-016395","kind":"patent","title":"A system managing the voting process","summary":"The voting system brings several AI methods into a collective decision process. The proposal addresses how participants communicate as a vote unfolds.","details":["Filed · Oct 31, 2022","Applicant · TÜRKİYE GARANTİ BANKASI ANONİM ŞİRKETİ","Published · Dec 21, 2022","TR2022016395A2","Portfolio-wide legal status is not provided by the dated extract. Individual grant decisions cite the national register."],"path":"/patents/2022-016395/","links":[{"label":"TR2022016395A2","url":"https://patents.google.com/patent/TR2022016395A2/en"},{"label":"Official TÜRKPATENT register","url":"https://www.turkpatent.gov.tr/arastirma-yap?form=patent"}],"meta":"TR 2022/016395 · Patent application","fields":["invention"],"year":"2022"},{"id":"model:turkish-sentiment","kind":"model","title":"Turkish Sentiment Analysis · TSAM","summary":"I trained this BERT model to read the tone of a Turkish review. It scores the sentence for positive, neutral and negative sentiment.","details":["AutoTrain classifier","An independent study tested four models on four Turkish datasets. In its binary evaluation, the researchers excluded neutral labels. This model scored 0.97 F1 on Humir and 0.54 on Twt.","Total Hub downloads · 12,053 · Nov 18, 2026"],"path":"/lab/turkish-sentiment/","links":[{"label":"Hugging Face","url":"https://huggingface.co/emre/turkish-sentiment-analysis"},{"label":"A cross-validation study of Turkish sentiment analysis datasets and tools","url":"https://doi.org/10.1007/s10579-025-09869-6"}],"meta":"Model · AutoTrain classifier","fields":["language"]},{"id":"model:whisper-tr","kind":"model","title":"Whisper Medium TR","summary":"Give it a Turkish recording and it returns a transcript. I adapted Whisper Medium on Mozilla Common Voice 11.0 speech to study recognition errors in Turkish.","details":["Speech-model adaptation","OpenAI Whisper Medium"],"path":"/lab/whisper-tr/","links":[{"label":"Hugging Face","url":"https://huggingface.co/emre/whisper-medium-turkish-2"},{"label":"OpenAI Whisper Medium","url":"https://huggingface.co/openai/whisper-medium"}],"meta":"Model · Speech-model adaptation","fields":["language"]},{"id":"model:gemma-wiki-tr","kind":"model","title":"Gemma 2 · Turkish Wikipedia","summary":"This adapter continues a Turkish passage using Gemma 2 9B trained further on Turkish Wikipedia. The card permits academic and research use and limits commercial use to BBVA Group.","details":["LoRA adapter author","Google Gemma 2 9B"],"path":"/lab/gemma-wiki-tr/","links":[{"label":"Hugging Face","url":"https://huggingface.co/emre/gemma-2-9b-Turkish-Lora-Continue-Pre-Trained"},{"label":"Google Gemma 2 9B","url":"https://huggingface.co/google/gemma-2-9b"}],"meta":"Model · LoRA adapter author","fields":["language"]},{"id":"model:java-method-vectors","kind":"model","title":"Java Methods · Doc2Vec","summary":"A Java method goes in and a 200-dimensional vector comes out. Developers can use these vectors to find similar snippets or group methods in a codebase.","details":["Embedding model developer","Gensim Doc2Vec"],"path":"/lab/java-method-vectors/","links":[{"label":"Hugging Face","url":"https://huggingface.co/emre/java_8m_methods_doc2vec"},{"label":"Gensim Doc2Vec","url":"https://radimrehurek.com/gensim/models/doc2vec.html"}],"meta":"Model · Embedding model developer","fields":["language"]},{"id":"model:gemma-reasoning-tr","kind":"model","title":"Gemma 3 27B · Turkish reasoning","summary":"I fine-tuned Unsloth’s 4-bit build of Gemma 3 27B Instruct for answering reasoning questions in Turkish. The card includes preliminary TARA evaluations on tasks such as logic, planning and SQL.","details":["Fine-tune of a quantized model","Unsloth 4-bit Gemma 3 27B Instruct"],"path":"/lab/gemma-reasoning-tr/","links":[{"label":"Hugging Face","url":"https://huggingface.co/emre/gemma-3-27b-it-tr-reasoning40k-4bit"},{"label":"Unsloth 4-bit Gemma 3 27B Instruct","url":"https://huggingface.co/unsloth/gemma-3-27b-it-unsloth-bnb-4bit"}],"meta":"Model · Fine-tune of a quantized model","fields":["language"]},{"id":"model:abstract-title","kind":"model","title":"arXiv Abstract-to-Title · mT5","summary":"Paste a research abstract to get a candidate paper title. This mT5-small adaptation learned the task from abstract–title pairs in an arXiv dataset.","details":["Sequence-to-sequence fine-tune","mT5-small"],"path":"/lab/abstract-title/","links":[{"label":"Hugging Face","url":"https://huggingface.co/emre/arxiv27k-t5-abst-title-gen"},{"label":"mT5-small","url":"https://huggingface.co/google/mt5-small"}],"meta":"Model · Sequence-to-sequence fine-tune","fields":["language"]},{"id":"model:xlsr-tr","kind":"model","title":"Wav2Vec2 XLS-R · Turkish","summary":"This Turkish speech recogniser turns recorded audio into text using the multilingual XLS-R encoder. Researchers can inspect its training history and its word error rate on Turkish Common Voice.","details":["Speech-model adaptation","Meta Wav2Vec2 XLS-R 300M"],"path":"/lab/xlsr-tr/","links":[{"label":"Hugging Face","url":"https://huggingface.co/emre/wav2vec2-large-xls-r-300m-tr"},{"label":"Meta Wav2Vec2 XLS-R 300M","url":"https://huggingface.co/facebook/wav2vec2-xls-r-300m"}],"meta":"Model · Speech-model adaptation","fields":["language"]},{"id":"model:qwen-grpo","kind":"model","title":"Qwen 0.5B · GRPO maths","summary":"A small model for maths word problems, trained on GSM8K to produce a worked response and a numerical answer. The training rewards answer correctness and response format separately.","details":["Reinforcement-learning fine-tune","Qwen2.5-0.5B-Instruct"],"path":"/lab/qwen-grpo/","links":[{"label":"Hugging Face","url":"https://huggingface.co/emre/Qwen-0.5B-GRPO"},{"label":"Qwen2.5-0.5B-Instruct","url":"https://huggingface.co/Qwen/Qwen2.5-0.5B-Instruct"}],"meta":"Model · Reinforcement-learning fine-tune","fields":["language"]},{"id":"repository:regression-atlas","kind":"repository","title":"Regression Conformal Prediction Research Atlas","summary":"Look up a dataset and a method to see how often its prediction intervals covered the outcome and how wide those intervals were. The atlas links results across 67 datasets to experiments and a browsable knowledge graph.","details":["Study author"],"path":"/lab/regression-atlas/","links":[{"label":"GitHub","url":"https://github.com/detasar/regression-conformal-prediction-research-atlas"}],"meta":"Repository · Study author","fields":["learning","confidence"],"year":"2026"},{"id":"repository:bird-sql","kind":"repository","title":"BIRD · Text-to-SQL experiments","summary":"The pipeline takes a question and a database schema, proposes SQL and checks the result by executing it. Researchers can inspect how ambiguity, table selection and query validation affect the answer on a financial subset of BIRD.","details":["Research repository maintainer"],"path":"/lab/bird-sql/","links":[{"label":"GitHub","url":"https://github.com/detasar/bird-text2sql-experiments"}],"meta":"Repository · Research repository maintainer","fields":["agents","confidence"]},{"id":"repository:webcli","kind":"repository","title":"WebCLI","summary":"WebCLI indexes public documentation and lets developers or agents search it locally from the command line. Results include stable references and can be exported as JSONL, with OpenAI’s developer documentation as the first adapter.","details":["Project maintainer"],"path":"/lab/webcli/","links":[{"label":"GitHub","url":"https://github.com/detasar/webcli"}],"meta":"Repository · Project maintainer","fields":["agents","language"]},{"id":"repository:fairness-metrics","kind":"repository","title":"Fairness Metrics","summary":"Give the Python library features, predictions, actual outcomes and a protected group. It returns a pandas table comparing quantities such as false-positive, false-negative and recall rates across groups.","details":["Library maintainer"],"path":"/lab/fairness-metrics/","links":[{"label":"GitHub","url":"https://github.com/detasar/fairness_metrics"}],"meta":"Repository · Library maintainer","fields":["learning","confidence"]},{"id":"repository:quantum-clocks","kind":"repository","title":"Quantum Time Experiment","summary":"If several fragments of a quantum system preserve the same record, can they tell us what happened earlier? The experiments test which records can be put in order and when an additional reference is needed to determine the direction.","details":["Research repository maintainer"],"path":"/lab/quantum-clocks/","links":[{"label":"GitHub","url":"https://github.com/detasar/quantum-time-experiment"}],"meta":"Repository · Research repository maintainer","fields":["quantum"]},{"id":"repository:paperbanana-test","kind":"repository","title":"PaperBanana field test","summary":"I tested Zhu and colleagues’ illustration pipeline on the methods sections of three of my papers. The repository pairs each untouched output with the original figure and exact input, so readers can judge where the illustrations went wrong.","details":["Independent evaluation author"],"path":"/lab/paperbanana-test/","links":[{"label":"GitHub","url":"https://github.com/detasar/paperbanana-field-test"}],"meta":"Repository · Independent evaluation author","fields":["agents"]},{"id":"repository:worldcup-nowcast","kind":"repository","title":"World Cup 2026 Nowcast Lab","summary":"This research lab turns match probabilities into distributions over possible tournament winners by simulating the 2026 World Cup. Readers can inspect the forecast card, data sources and model choices behind each snapshot.","details":["Forecasting project maintainer"],"path":"/lab/worldcup-nowcast/","links":[{"label":"GitHub","url":"https://github.com/detasar/worldcup26-nowcast-lab"}],"meta":"Repository · Forecasting project maintainer","fields":["learning","confidence"]},{"id":"repository:agent-chronicle","kind":"repository","title":"Agent Chronicle","summary":"Read a Codex or Claude session as a timeline of decisions, changes and checks. Each recorded event gives the developer a way to inspect the evidence behind a progress report.","details":["Project maintainer"],"path":"/lab/agent-chronicle/","links":[{"label":"GitHub","url":"https://github.com/detasar/agent-chronicle-community"}],"meta":"Repository · Project maintainer","fields":["agents"]},{"id":"repository:turkish-binary-reviews","kind":"repository","title":"BERT-TR-Finetuned4-Pos-Neg-Reviews","summary":"I fine-tuned Turkish BERT to tell positive reviews from negative ones. The training settings are on GitHub.","details":["An earlier review classifier","dbmdz/bert-base-turkish-cased"],"path":"/lab/turkish-sentiment/","links":[{"label":"Repository","url":"https://github.com/detasar/BERT-TR-Finetuned4-Pos-Neg-Reviews"},{"label":"Original source","url":"https://github.com/detasar/BERT-TR-Finetuned4-Pos-Neg-Reviews/blob/7e6c1d1065063e51872026bce286a0fe1c74ebf8/model_args.json"}],"fields":["language","learning"],"year":"2021"},{"id":"study:coverage-deferral","kind":"study","title":"Table 4 · The coverage–deferral trade-off","summary":"Whose case gets a second look?","details":["100 random seeds. Target coverage: 90%.","Percentage points","Adult · Coverage gap (Pooled calibration 5 / Groupwise 1.2) · Referral gap (Pooled calibration 5.1 / Groupwise 12.9)","ACS Income · Coverage gap (Pooled calibration 0.8 / Groupwise 0.8) · Referral gap (Pooled calibration 3.4 / Groupwise 5.4)","Bank Marketing · Coverage gap (Pooled calibration 2 / Groupwise 1.1) · Referral gap (Pooled calibration 0 / Groupwise 3)","COMPAS · Coverage gap (Pooled calibration 1.8 / Groupwise 3.3) · Referral gap (Pooled calibration 5.7 / Groupwise 10.8)","German Credit · Coverage gap (Pooled calibration 3.5 / Groupwise 5.7) · Referral gap (Pooled calibration 6.1 / Groupwise 24.7)","Taiwan Credit · Coverage gap (Pooled calibration 1.1 / Groupwise 1.3) · Referral gap (Pooled calibration 3.4 / Groupwise 7.7)"],"path":"/#study","links":[{"label":"Table 4","url":"https://www.preprints.org/manuscript/202512.2631#:~:text=Table%204.%20Coverage-deferral%20trade-off%20across%20datasets"},{"label":"Original source","url":"https://www.preprints.org/manuscript/202512.2631"}],"fields":["confidence","learning"]},{"id":"study:turkish-sentiment-evaluation","kind":"study","title":"A cross-validation study of Turkish sentiment analysis datasets and tools","summary":"An independent study tested four models on four Turkish datasets. In its binary evaluation, the researchers excluded neutral labels. This model scored 0.97 F1 on Humir and 0.54 on Twt.","details":["Humir · F1 0.97","Twt · F1 0.54"],"path":"/lab/turkish-sentiment/","links":[{"label":"Original source","url":"https://doi.org/10.1007/s10579-025-09869-6"}],"fields":["language"]},{"id":"contact:github","kind":"contact","title":"GitHub","summary":"Code repositories","details":["Davut Emre Taşar"],"path":"/#profiles","links":[{"label":"GitHub","url":"https://github.com/detasar"}],"fields":[]},{"id":"contact:hugging-face","kind":"contact","title":"Hugging Face","summary":"Models","details":["Davut Emre Taşar"],"path":"/#profiles","links":[{"label":"Hugging Face","url":"https://huggingface.co/emre"}],"fields":[]},{"id":"contact:scholar","kind":"contact","title":"Google Scholar","summary":"Research and publications","details":["Davut Emre Taşar"],"path":"/#profiles","links":[{"label":"Google Scholar","url":"https://scholar.google.com/citations?user=IZZM_JAAAAAJ"}],"fields":[]},{"id":"contact:linkedin","kind":"contact","title":"LinkedIn","summary":"Contact me on LinkedIn","details":["Davut Emre Taşar"],"path":"/#profiles","links":[{"label":"LinkedIn","url":"https://www.linkedin.com/in/detasar/"}],"fields":[]},{"id":"contact:x","kind":"contact","title":"X","summary":"@mehanixs","details":["Davut Emre Taşar"],"path":"/#profiles","links":[{"label":"X","url":"https://x.com/mehanixs"}],"fields":[]},{"id":"contact:orcid","kind":"contact","title":"ORCID","summary":"ORCID · 0000-0002-7788-0478","details":["Davut Emre Taşar"],"path":"/#profiles","links":[{"label":"ORCID","url":"https://orcid.org/0000-0002-7788-0478"}],"fields":[]},{"id":"contact:navarra","kind":"contact","title":"University of Navarra","summary":"Doctoral researcher in AI fairness","details":["Davut Emre Taşar"],"path":"/#profiles","links":[{"label":"University of Navarra","url":"https://www.unav.edu/web/instituto-de-ciencia-de-los-datos-e-inteligencia-artificial/equipo#:~:text=Enhancing%20Fairness%20in%20AI%20Systems"}],"fields":[]},{"id":"reader:guide","kind":"reader","title":"Reading index","summary":"Search research, Turkish patent and utility model applications, models and code repositories in one place.","details":["Index, graph and profile"],"path":"/guide/","links":[{"label":"llms.txt","url":"/llms.txt"}],"fields":[]},{"id":"reader:graph","kind":"reader","title":"Knowledge graph","summary":"Choose a research area to explore how the work connects and follow the links to its original sources.","details":["Index, graph and profile"],"path":"/graph/","links":[{"label":"ontology.jsonld","url":"/ontology.jsonld"}],"fields":[]},{"id":"reader:profile","kind":"reader","title":"Profile","summary":"I’m an engineer and applied scientist on BBVA’s AI Innovation team, part of AI Transformation. I’m doing a PhD at the University of Navarra on fairness in AI.","details":["Index, graph and profile"],"path":"/profile/","links":[{"label":"profile.json","url":"/profile.json"}],"fields":[]}],"edges":[{"from":"profile:emre","to":"section:selected","type":"related"},{"from":"profile:emre","to":"section:research","type":"related"},{"from":"profile:emre","to":"section:works","type":"related"},{"from":"profile:emre","to":"section:notebook","type":"related"},{"from":"profile:emre","to":"section:portfolio","type":"related"},{"from":"profile:emre","to":"section:lab","type":"related"},{"from":"profile:emre","to":"section:album","type":"related"},{"from":"profile:emre","to":"section:contact","type":"related"},{"from":"profile:emre","to":"section:readers","type":"related"},{"from":"section:selected","to":"project:meaning-native-computing","type":"contains"},{"from":"section:selected","to":"work:micro-robot-rescue","type":"contains"},{"from":"section:selected","to":"work:quantum-conformal","type":"contains"},{"from":"section:selected","to":"work:weight-editing","type":"contains"},{"from":"section:research","to":"topic:uncertainty","type":"contains"},{"from":"topic:uncertainty","to":"topic:fairness","type":"related"},{"from":"topic:uncertainty","to":"topic:quantum","type":"related"},{"from":"topic:uncertainty","to":"topic:agents","type":"related"},{"from":"section:research","to":"topic:fairness","type":"contains"},{"from":"topic:fairness","to":"topic:agents","type":"related"},{"from":"topic:fairness","to":"topic:invention","type":"related"},{"from":"section:research","to":"topic:agents","type":"contains"},{"from":"topic:agents","to":"topic:invention","type":"related"},{"from":"topic:agents","to":"topic:language","type":"related"},{"from":"section:research","to":"topic:quantum","type":"contains"},{"from":"topic:quantum","to":"topic:fairness","type":"related"},{"from":"section:research","to":"topic:invention","type":"contains"},{"from":"topic:invention","to":"topic:language","type":"related"},{"from":"section:research","to":"topic:language","type":"contains"},{"from":"topic:language","to":"topic:uncertainty","type":"related"},{"from":"profile:emre","to":"affiliation:bbva","type":"affiliation"},{"from":"profile:emre","to":"affiliation:navarra","type":"affiliation"},{"from":"section:album","to":"photo:espresso","type":"contains"},{"from":"section:album","to":"photo:scholar","type":"contains"},{"from":"photo:scholar","to":"affiliation:navarra","type":"related"},{"from":"section:album","to":"photo:we-think-big","type":"contains"},{"from":"photo:we-think-big","to":"affiliation:bbva","type":"related"},{"from":"section:notebook","to":"note:ontologies-and-data-structures","type":"contains"},{"from":"section:notebook","to":"note:category-theory","type":"contains"},{"from":"section:notebook","to":"note:quantum-measurement","type":"contains"},{"from":"note:quantum-measurement","to":"topic:quantum","type":"about"},{"from":"section:notebook","to":"project:meaning-native-computing","type":"contains"},{"from":"profile:emre","to":"project:meaning-native-computing","type":"related"},{"from":"project:meaning-native-computing","to":"topic:uncertainty","type":"about"},{"from":"section:works","to":"work:coverage-deferral","type":"contains"},{"from":"work:coverage-deferral","to":"topic:fairness","type":"about"},{"from":"work:coverage-deferral","to":"topic:uncertainty","type":"about"},{"from":"section:works","to":"work:agent-chronicle","type":"contains"},{"from":"work:agent-chronicle","to":"topic:agents","type":"about"},{"from":"section:works","to":"work:quantum-conformal","type":"contains"},{"from":"work:quantum-conformal","to":"topic:uncertainty","type":"about"},{"from":"work:quantum-conformal","to":"topic:quantum","type":"about"},{"from":"section:works","to":"work:tara","type":"contains"},{"from":"work:tara","to":"topic:quantum","type":"about"},{"from":"work:tara","to":"topic:uncertainty","type":"about"},{"from":"section:works","to":"work:chat-assistant","type":"contains"},{"from":"work:chat-assistant","to":"topic:invention","type":"about"},{"from":"work:chat-assistant","to":"topic:agents","type":"about"},{"from":"work:chat-assistant","to":"topic:language","type":"about"},{"from":"section:works","to":"work:micro-robot-rescue","type":"contains"},{"from":"work:micro-robot-rescue","to":"topic:invention","type":"about"},{"from":"work:micro-robot-rescue","to":"topic:agents","type":"about"},{"from":"section:works","to":"work:ai-pos","type":"contains"},{"from":"work:ai-pos","to":"topic:invention","type":"about"},{"from":"work:ai-pos","to":"topic:language","type":"about"},{"from":"section:works","to":"work:virtual-input","type":"contains"},{"from":"work:virtual-input","to":"topic:invention","type":"about"},{"from":"section:works","to":"work:personalized-suggestions","type":"contains"},{"from":"work:personalized-suggestions","to":"topic:invention","type":"about"},{"from":"section:works","to":"work:adversarial-quantum","type":"contains"},{"from":"work:adversarial-quantum","to":"topic:quantum","type":"about"},{"from":"work:adversarial-quantum","to":"topic:uncertainty","type":"about"},{"from":"section:works","to":"work:turkish-asr","type":"contains"},{"from":"work:turkish-asr","to":"topic:language","type":"about"},{"from":"section:works","to":"work:qwen-grpo","type":"contains"},{"from":"work:qwen-grpo","to":"topic:language","type":"about"},{"from":"section:works","to":"work:turkish-reasoning-benchmark","type":"contains"},{"from":"work:turkish-reasoning-benchmark","to":"topic:language","type":"about"},{"from":"section:works","to":"work:multilingual-chatbots","type":"contains"},{"from":"work:multilingual-chatbots","to":"topic:language","type":"about"},{"from":"section:works","to":"work:ad-text-classification","type":"contains"},{"from":"work:ad-text-classification","to":"topic:language","type":"about"},{"from":"section:works","to":"work:weight-editing","type":"contains"},{"from":"work:weight-editing","to":"topic:agents","type":"about"},{"from":"work:weight-editing","to":"topic:language","type":"about"},{"from":"section:works","to":"work:bloom","type":"contains"},{"from":"work:bloom","to":"topic:language","type":"about"},{"from":"section:portfolio","to":"patent:2026-007909","type":"contains"},{"from":"patent:2026-007909","to":"topic:invention","type":"about"},{"from":"patent:2026-007909","to":"topic:language","type":"about"},{"from":"section:portfolio","to":"patent:2026-007907","type":"contains"},{"from":"patent:2026-007907","to":"topic:invention","type":"about"},{"from":"section:portfolio","to":"patent:2026-006294","type":"contains"},{"from":"patent:2026-006294","to":"topic:invention","type":"about"},{"from":"section:portfolio","to":"patent:2026-006289","type":"contains"},{"from":"patent:2026-006289","to":"topic:invention","type":"about"},{"from":"section:portfolio","to":"patent:2026-004444","type":"contains"},{"from":"patent:2026-004444","to":"topic:invention","type":"about"},{"from":"section:portfolio","to":"patent:2026-004441","type":"contains"},{"from":"patent:2026-004441","to":"topic:invention","type":"about"},{"from":"patent:2026-004441","to":"topic:quantum","type":"about"},{"from":"patent:2026-004441","to":"topic:uncertainty","type":"about"},{"from":"section:portfolio","to":"patent:2026-002909","type":"contains"},{"from":"patent:2026-002909","to":"topic:invention","type":"about"},{"from":"section:portfolio","to":"patent:2026-001992","type":"contains"},{"from":"patent:2026-001992","to":"topic:invention","type":"about"},{"from":"section:portfolio","to":"patent:2025-019989","type":"contains"},{"from":"patent:2025-019989","to":"topic:invention","type":"about"},{"from":"section:portfolio","to":"patent:2025-011133","type":"contains"},{"from":"patent:2025-011133","to":"topic:invention","type":"about"},{"from":"section:portfolio","to":"patent:2025-008786","type":"contains"},{"from":"patent:2025-008786","to":"topic:invention","type":"about"},{"from":"section:portfolio","to":"patent:2025-008273","type":"contains"},{"from":"patent:2025-008273","to":"topic:invention","type":"about"},{"from":"section:portfolio","to":"patent:2025-008055","type":"contains"},{"from":"patent:2025-008055","to":"topic:invention","type":"about"},{"from":"section:portfolio","to":"patent:2025-008032","type":"contains"},{"from":"patent:2025-008032","to":"topic:invention","type":"about"},{"from":"section:portfolio","to":"patent:2025-008028","type":"contains"},{"from":"patent:2025-008028","to":"topic:invention","type":"about"},{"from":"section:portfolio","to":"patent:2025-007997","type":"contains"},{"from":"patent:2025-007997","to":"topic:invention","type":"about"},{"from":"section:portfolio","to":"patent:2025-007995","type":"contains"},{"from":"patent:2025-007995","to":"topic:invention","type":"about"},{"from":"patent:2025-007995","to":"topic:agents","type":"about"},{"from":"section:portfolio","to":"patent:2024-003815","type":"contains"},{"from":"patent:2024-003815","to":"topic:invention","type":"about"},{"from":"section:portfolio","to":"patent:2024-003813","type":"contains"},{"from":"patent:2024-003813","to":"topic:invention","type":"about"},{"from":"section:portfolio","to":"patent:2024-003592","type":"contains"},{"from":"patent:2024-003592","to":"topic:invention","type":"about"},{"from":"section:portfolio","to":"patent:2023-011901","type":"contains"},{"from":"patent:2023-011901","to":"topic:invention","type":"about"},{"from":"patent:2023-011901","to":"work:micro-robot-rescue","type":"family"},{"from":"section:portfolio","to":"patent:2023-011870","type":"contains"},{"from":"patent:2023-011870","to":"topic:invention","type":"about"},{"from":"patent:2023-011870","to":"topic:language","type":"about"},{"from":"patent:2023-011870","to":"work:chat-assistant","type":"family"},{"from":"section:portfolio","to":"patent:2023-011830","type":"contains"},{"from":"patent:2023-011830","to":"topic:invention","type":"about"},{"from":"patent:2023-011830","to":"work:personalized-suggestions","type":"family"},{"from":"section:portfolio","to":"patent:2023-011601","type":"contains"},{"from":"patent:2023-011601","to":"topic:invention","type":"about"},{"from":"patent:2023-011601","to":"work:virtual-input","type":"family"},{"from":"section:portfolio","to":"patent:2023-011600","type":"contains"},{"from":"patent:2023-011600","to":"topic:invention","type":"about"},{"from":"section:portfolio","to":"patent:2023-011599","type":"contains"},{"from":"patent:2023-011599","to":"topic:invention","type":"about"},{"from":"section:portfolio","to":"patent:2023-004244","type":"contains"},{"from":"patent:2023-004244","to":"topic:invention","type":"about"},{"from":"patent:2023-004244","to":"work:ai-pos","type":"family"},{"from":"section:portfolio","to":"patent:2023-004047","type":"contains"},{"from":"patent:2023-004047","to":"topic:invention","type":"about"},{"from":"section:portfolio","to":"patent:2022-016679","type":"contains"},{"from":"patent:2022-016679","to":"topic:invention","type":"about"},{"from":"section:portfolio","to":"patent:2022-016395","type":"contains"},{"from":"patent:2022-016395","to":"topic:invention","type":"about"},{"from":"section:lab","to":"model:turkish-sentiment","type":"contains"},{"from":"model:turkish-sentiment","to":"topic:language","type":"about"},{"from":"section:lab","to":"model:whisper-tr","type":"contains"},{"from":"model:whisper-tr","to":"topic:language","type":"about"},{"from":"section:lab","to":"model:gemma-wiki-tr","type":"contains"},{"from":"model:gemma-wiki-tr","to":"topic:language","type":"about"},{"from":"section:lab","to":"model:java-method-vectors","type":"contains"},{"from":"model:java-method-vectors","to":"topic:language","type":"about"},{"from":"section:lab","to":"model:gemma-reasoning-tr","type":"contains"},{"from":"model:gemma-reasoning-tr","to":"topic:language","type":"about"},{"from":"section:lab","to":"model:abstract-title","type":"contains"},{"from":"model:abstract-title","to":"topic:language","type":"about"},{"from":"section:lab","to":"model:xlsr-tr","type":"contains"},{"from":"model:xlsr-tr","to":"topic:language","type":"about"},{"from":"section:lab","to":"model:qwen-grpo","type":"contains"},{"from":"model:qwen-grpo","to":"topic:language","type":"about"},{"from":"work:qwen-grpo","to":"model:qwen-grpo","type":"sameProject"},{"from":"section:lab","to":"repository:regression-atlas","type":"contains"},{"from":"repository:regression-atlas","to":"topic:uncertainty","type":"about"},{"from":"section:lab","to":"repository:bird-sql","type":"contains"},{"from":"repository:bird-sql","to":"topic:agents","type":"about"},{"from":"repository:bird-sql","to":"topic:uncertainty","type":"about"},{"from":"section:lab","to":"repository:webcli","type":"contains"},{"from":"repository:webcli","to":"topic:agents","type":"about"},{"from":"repository:webcli","to":"topic:language","type":"about"},{"from":"section:lab","to":"repository:fairness-metrics","type":"contains"},{"from":"repository:fairness-metrics","to":"topic:fairness","type":"about"},{"from":"section:lab","to":"repository:quantum-clocks","type":"contains"},{"from":"repository:quantum-clocks","to":"topic:quantum","type":"about"},{"from":"section:lab","to":"repository:paperbanana-test","type":"contains"},{"from":"repository:paperbanana-test","to":"topic:agents","type":"about"},{"from":"section:lab","to":"repository:worldcup-nowcast","type":"contains"},{"from":"repository:worldcup-nowcast","to":"topic:uncertainty","type":"about"},{"from":"section:lab","to":"repository:agent-chronicle","type":"contains"},{"from":"repository:agent-chronicle","to":"topic:agents","type":"about"},{"from":"work:agent-chronicle","to":"repository:agent-chronicle","type":"sameProject"},{"from":"model:turkish-sentiment","to":"repository:turkish-binary-reviews","type":"related"},{"from":"repository:turkish-binary-reviews","to":"topic:language","type":"about"},{"from":"study:coverage-deferral","to":"work:coverage-deferral","type":"documents"},{"from":"study:coverage-deferral","to":"topic:fairness","type":"about"},{"from":"study:coverage-deferral","to":"topic:uncertainty","type":"about"},{"from":"study:turkish-sentiment-evaluation","to":"model:turkish-sentiment","type":"documents"},{"from":"study:turkish-sentiment-evaluation","to":"topic:language","type":"about"},{"from":"section:contact","to":"contact:github","type":"contains"},{"from":"profile:emre","to":"contact:github","type":"contact"},{"from":"contact:github","to":"repository:regression-atlas","type":"related"},{"from":"contact:github","to":"repository:bird-sql","type":"related"},{"from":"contact:github","to":"repository:webcli","type":"related"},{"from":"contact:github","to":"repository:fairness-metrics","type":"related"},{"from":"contact:github","to":"repository:quantum-clocks","type":"related"},{"from":"contact:github","to":"repository:paperbanana-test","type":"related"},{"from":"contact:github","to":"repository:worldcup-nowcast","type":"related"},{"from":"contact:github","to":"repository:agent-chronicle","type":"related"},{"from":"section:contact","to":"contact:hugging-face","type":"contains"},{"from":"profile:emre","to":"contact:hugging-face","type":"contact"},{"from":"contact:hugging-face","to":"model:turkish-sentiment","type":"related"},{"from":"contact:hugging-face","to":"model:whisper-tr","type":"related"},{"from":"contact:hugging-face","to":"model:gemma-wiki-tr","type":"related"},{"from":"contact:hugging-face","to":"model:java-method-vectors","type":"related"},{"from":"contact:hugging-face","to":"model:gemma-reasoning-tr","type":"related"},{"from":"contact:hugging-face","to":"model:abstract-title","type":"related"},{"from":"contact:hugging-face","to":"model:xlsr-tr","type":"related"},{"from":"contact:hugging-face","to":"model:qwen-grpo","type":"related"},{"from":"section:contact","to":"contact:scholar","type":"contains"},{"from":"profile:emre","to":"contact:scholar","type":"contact"},{"from":"section:contact","to":"contact:linkedin","type":"contains"},{"from":"profile:emre","to":"contact:linkedin","type":"contact"},{"from":"affiliation:bbva","to":"contact:linkedin","type":"documents"},{"from":"section:contact","to":"contact:x","type":"contains"},{"from":"profile:emre","to":"contact:x","type":"contact"},{"from":"section:contact","to":"contact:orcid","type":"contains"},{"from":"profile:emre","to":"contact:orcid","type":"contact"},{"from":"section:contact","to":"contact:navarra","type":"contains"},{"from":"profile:emre","to":"contact:navarra","type":"contact"},{"from":"affiliation:navarra","to":"contact:navarra","type":"documents"},{"from":"section:readers","to":"reader:guide","type":"contains"},{"from":"section:readers","to":"reader:graph","type":"contains"},{"from":"section:readers","to":"reader:profile","type":"contains"},{"from":"profile:emre","to":"reader:profile","type":"readableAs"},{"from":"section:research","to":"reader:graph","type":"readableAs"},{"from":"section:portfolio","to":"reader:guide","type":"readableAs"},{"from":"section:lab","to":"reader:guide","type":"readableAs"},{"from":"section:works","to":"reader:guide","type":"readableAs"}]}