BEGIN:VCALENDAR PRODID:-//github.com/ical-org/ical.net//NONSGML ical.net 4.0//EN VERSION:2.0 BEGIN:VTIMEZONE TZID:America/New_York X-LIC-LOCATION:America/New_York BEGIN:STANDARD DTSTART:20241103T020000 RRULE:FREQ=YEARLY;BYDAY=1SU;BYMONTH=11 TZNAME:EST TZOFFSETFROM:-0400 TZOFFSETTO:-0500 END:STANDARD BEGIN:DAYLIGHT DTSTART:20250309T020000 RRULE:FREQ=YEARLY;BYDAY=2SU;BYMONTH=3 TZNAME:EDT TZOFFSETFROM:-0500 TZOFFSETTO:-0400 END:DAYLIGHT END:VTIMEZONE BEGIN:VEVENT DESCRIPTION:Join us for the next seminar in our NLP/LLM Interest Group\, f eaturing cutting-edge research at the intersection of language models\, e mbeddings\, and multimodal learning. Featured Talks: A Vector is Worth 1\ ,000 Words: Training Large Language Models to Interpret Embedding Spaces Speaker: Brian Ondov\, PhD\, Associate Research Scientist Empowering Mult imodal Large Language Models for Grounded ECG Understanding with Time Ser ies and Images Speaker: Xiang Lan\, PhD\, Postdoctoral Associate ⬅️ Downl oad the flyer for abstract details (PDF) 💌 Subscribe to our mailing list to stay informed about future events.\n\nSpeakers:\nBrian Ondov\; Xiang Lan\n\nAdmission:\nFree\n\nDetails URL:\nhttps://medicine.yale.edu/event/ nlpllm-interest-group-1/\n DTEND;TZID=America/New_York:20250915T170000 DTSTAMP:20260429T090341Z DTSTART;TZID=America/New_York:20250915T160000 LOCATION:URL: https://yale.zoom.us/j/93599941969 RECURRENCE-ID;TZID=America/New_York:20250915T160000 SEQUENCE:0 STATUS:Confirmed SUMMARY:NLP/LLM Interest Group Session UID:e776ef46-f30b-4de3-80b8-65f255e3ba90 END:VEVENT BEGIN:VEVENT DESCRIPTION:Danielle Mowery\, PhD\, is a collaborative investigator that d evelops natural language processing (NLP) and generative artificial intel ligence (AI) solutions for processing clinical texts – i.e.\, clinical no tes\, chatbots\, and transcribed texts – to support clinical and translat ional research. She leverages NLP\, data science\, machine learning\, and computational methods to integrate and analyze information from unstruct ured texts and structured clinical data to help clinical investigators be tter understand disease burden\, treatment efficacy\, and clinical outcom es. Furthermore\, her solutions focus on helping patients and clinicians make better decisions at the point of care whether it’s in a traditional health system setting or through digital health services within a patient ’s home. Her work aims to uncover scientific discoveries\, identify actio nable healthcare knowledge\, and optimize translation of research into pa tient care. In this talk\, she will review key use cases in which NLP and generative AI are innovating in key areas of basic science\, applied cli nical informatics\, and population health.\n\nSpeaker:\nDanielle L. Mower y\, PhD\, MS\, MS\, FAMIA \n\nAdmission:\nFree\n\nDetails URL:\nhttps://m edicine.yale.edu/event/nlpllm-interest-group-mowery/\n DTEND;TZID=America/New_York:20250922T170000 DTSTAMP:20260429T090341Z DTSTART;TZID=America/New_York:20250922T160000 LOCATION:URL: https://yale.zoom.us/j/93599941969 RECURRENCE-ID;TZID=America/New_York:20250922T160000 SEQUENCE:0 STATUS:Confirmed SUMMARY:NLP/LLM Interest Group UID:e776ef46-f30b-4de3-80b8-65f255e3ba90 END:VEVENT BEGIN:VEVENT DESCRIPTION:Andrew Loza\, MD\, PhD\, will introduce Comet—a family of tran sformer models trained on Epic Cosmos\, an unprecedented dataset containi ng 16.3 billion encounters across 300 million patient records from 310 he alth systems\, representing 115 billion medical events from 118 million p atients. Comet autoregressively predicts the next medical event to simula te patient health timelines. Across 78 real-world tasks including diagnos is prediction\, disease prognosis\, and healthcare operations\, Comet out performed or matched task-specific supervised models without requiring fi ne-tuning. Our results demonstrate that generative medical event foundati on models can effectively capture complex clinical dynamics\, providing a generalizable framework to support clinical decision-making and improve patient outcomes.\n\nSpeaker:\nAndrew Loza\n\nAdmission:\nFree\n\nDetails URL:\nhttps://medicine.yale.edu/event/nlpllm-interest-group-3/\n DTEND;TZID=America/New_York:20250929T170000 DTSTAMP:20260429T090341Z DTSTART;TZID=America/New_York:20250929T160000 LOCATION:Zoom - Passcode Required\, URL: https://yale.zoom.us/j/93599941969 RECURRENCE-ID;TZID=America/New_York:20250929T160000 SEQUENCE:0 STATUS:Confirmed SUMMARY:NLP/LLM Interest Group UID:e776ef46-f30b-4de3-80b8-65f255e3ba90 END:VEVENT BEGIN:VEVENT DESCRIPTION:The rapid growth of scientific publications has made literatur e discovery increasingly challenging. While traditional search engines an d recent chatbots provide access and summaries\, they remain limited for deeper exploration. In this talk\, Huan He\, PhD will introduce the conce pt of AI agents as "co-pilots" for literature discovery\, using the MedVi z system as a case study. He will demonstrate how multi-agent architectur es and large-scale visualizations can transform literature search from a passive query-response model into an active\, exploratory process.\n\nSpe aker:\nHuan He\n\nAdmission:\nFree\n\nDetails URL:\nhttps://medicine.yale .edu/event/nlpllm-interest-group-4/\n DTEND;TZID=America/New_York:20251006T170000 DTSTAMP:20260429T090341Z DTSTART;TZID=America/New_York:20251006T160000 LOCATION:Passcode Required: 849811\, URL: https://yale.zoom.us/j/935999419 69 RECURRENCE-ID;TZID=America/New_York:20251006T160000 SEQUENCE:0 STATUS:Confirmed SUMMARY:NLP/LLM Interest Group UID:e776ef46-f30b-4de3-80b8-65f255e3ba90 END:VEVENT BEGIN:VEVENT DESCRIPTION:\nAdmission:\nFree\n\nDetails URL:\nhttps://medicine.yale.edu/ event/nlpllm-interest-group-5/\n DTEND;TZID=America/New_York:20251013T170000 DTSTAMP:20260429T090341Z DTSTART;TZID=America/New_York:20251013T160000 LOCATION:Zoom: Passcode Required\, URL: https://yale.zoom.us/j/93599941969 RECURRENCE-ID;TZID=America/New_York:20251013T160000 SEQUENCE:0 STATUS:Cancelled SUMMARY:NLP/LLM Interest Group UID:e776ef46-f30b-4de3-80b8-65f255e3ba90 END:VEVENT BEGIN:VEVENT DESCRIPTION:The evidence base for LLMs in digestive diseases shows wide pe rformance variability\, underscoring safety risks and the need for rigoro us evaluation. In this talk\, Mauro Giuffrè\, MD will overview his main c ontributions in the field: a systematic review that quantified accuracy r anges and highlighted methodological gaps\; a guideline-grounded study sh owing that retrieval-augmented and fine-tuned GPT-4 markedly improve open -ended answer quality and treatment selection in patients with Hepatitis C Virus\; an “expert-of-experts” verification framework (EVAL) that align s automated grading with human experts and boosts correctness via rejecti on sampling\; and a randomized simulation trial (GutGPT) revealing that b etter usability does not automatically translate into adoption\, pointing to trust and workflow integration as key levers for impact.\n\nAdmission :\nFree\n\nDetails URL:\nhttps://medicine.yale.edu/event/nlpllm-interest- group-giuffre/\n DTEND;TZID=America/New_York:20251020T170000 DTSTAMP:20260429T090341Z DTSTART;TZID=America/New_York:20251020T160000 LOCATION:Zoom: Passcode Required: 849811 \, URL: https://yale.zoom.us/j/93 599941969 RECURRENCE-ID;TZID=America/New_York:20251020T160000 SEQUENCE:0 STATUS:Confirmed SUMMARY:NLP/LLM Interest Group UID:e776ef46-f30b-4de3-80b8-65f255e3ba90 END:VEVENT BEGIN:VEVENT DESCRIPTION:From Compound Figures to Composite Understanding: Developing a Multi-Modal LLM from Biomedical Literature with Medical Multiple-Image B enchmarking and Validation In healthcare\, disease diagnostics and longit udinal patient monitoring require clinicians to synthesize information ac ross multiple images from different modalities or time points\, yet this multi-image reasoning remains a significant gap for most current multi-mo dal LLMs. This capability gap persists due to a critical bottleneck: the lack of large-scale\, high-quality annotated training data for medical mu lti-image understanding. This study aims to address this scarcity by leve raging compound figures from biomedical literature. We devised a novel fi ve-stage\, context-aware instruction generation pipeline to create the PM C-MI-Dataset comprising over 260\,000 compound images\, and subsequently developed M³LLM\, a medical multi-image multi-modal LLM. For a comprehens ive evaluation\, we also constructed the expert-validated PMC-MI-Bench. M ³LLM significantly outperforms state-of-the-art general-purpose and speci alized MLLMs\, achieving superior performance on diverse tasks of the PMC -MI-Bench and public benchmarks like OmniMedVQA and MMMU-Med. Furthermore \, clinical validation on the MIMIC longitudinal chest X-ray dataset conf irms its superior performance in real-world tasks\, including disease dia gnosis and progression prediction. Our study establishes a scalable parad igm for this task\, and the model\, dataset\, and benchmark will be publi cly released.\n\nAdmission:\nFree\n\nDetails URL:\nhttps://medicine.yale. edu/event/nlpllm-interest-group-7/\n DTEND;TZID=America/New_York:20251027T170000 DTSTAMP:20260429T090341Z DTSTART;TZID=America/New_York:20251027T160000 LOCATION:Zoom: Passcode Required\, URL: https://yale.zoom.us/j/93599941969 RECURRENCE-ID;TZID=America/New_York:20251027T160000 SEQUENCE:0 STATUS:Confirmed SUMMARY:NLP/LLM Interest Group UID:e776ef46-f30b-4de3-80b8-65f255e3ba90 END:VEVENT BEGIN:VEVENT DESCRIPTION:Large language models hold enormous potential for transforming healthcare\, but their deployment is often bottlenecked by unstructured data\, fragmented pipelines\, and inconsistent model-data alignment. Nimb lemind.AI addresses these challenges with an agentic AI framework that au tomates the full clinical data lifecycle from ingestion and de-identifica tion to inference and interpretability. Building on the paper\, “Agentic AI Framework for End-to-End Medical Data Inference”\, the talk will discu ss how specialized agents collaborate to process multimodal data\, link s tructured EHR fields with clinical narratives\, and route tasks to the mo st appropriate foundation or domain model. This framework demonstrates ho w autonomous agent coordination can turn LLMs into operational tools to c ontextualize patient records\, ensure reproducibility\, and generate tran sparent explanations of model outputs. The talk will wrap up with a demo of Nimblemind.AI's new platform\, NimbleLabs\, where users can add a simp le clinical description and relevant data to generate a fully operational \, specialty-specific predictive model ready for real-world clinical envi ronments. Navin Kumar\, PhD is the co-founder of Nimblemind.AI\, a compan y building agentic AI systems to make clinical data interoperable and act ionable. Born and raised in Singapore\, Kumar earned their PhD at Yale Un iversity under Nicholas Christakis \, publishing over 60 papers and secur ing more than $1M in research funding on AI-driven approaches to populati on health. Before founding Nimblemind.AI\, they led data science and AI i nitiatives at NYC Health + Hospitals\, the nation’s largest public health care system\, and deployed predictive models that improved chronic diseas e screening and reduced missed appointments across over one million patie nts. Witnessing firsthand how data inefficiency limits equitable care\, K umar launched Nimblemind.AI to build scalable\, transparent AI infrastruc ture that improves outcomes\, especially for minority populations.\n\nSpe aker:\nNavin Kumar\, PhD\n\nAdmission:\nFree\n\nDetails URL:\nhttps://med icine.yale.edu/event/nlpllm-interest-group-8/\n DTEND;TZID=America/New_York:20251103T170000 DTSTAMP:20260429T090341Z DTSTART;TZID=America/New_York:20251103T160000 LOCATION:Zoom: Passcode Required\, URL: https://yale.zoom.us/j/93599941969 RECURRENCE-ID;TZID=America/New_York:20251103T160000 SEQUENCE:0 STATUS:Confirmed SUMMARY:NLP/LLM Interest Group UID:e776ef46-f30b-4de3-80b8-65f255e3ba90 END:VEVENT BEGIN:VEVENT DESCRIPTION:Talk Title: RDMA: Cost Effective Agent-Driven Rare Disease Dis covery within Electronic Health Record Systems Abstract: Rare diseases af fect 1 in 10 Americans\, yet standard ICD coding systems fail to capture these conditions in electronic health records (EHR)\, leaving crucial inf ormation about rare diseases\, their clinical presentations\, and phenoty pic patterns buried in unstructured clinical notes. Current automated ext raction approaches struggle with medical abbreviations\, miss implicit ph enotype mentions\, raise privacy concerns through cloud processing\, and lack the clinical reasoning abilities needed for accurate identification of rare disease presentations in human patients. We present Rare Disease Mining Agents (RDMA)\, a framework that mirrors how clinical experts appr oach rare disease identification by systematically connecting clinical ob servations directly to standardized ontologies like Orphanet and Human Ph enotype Ontology. RDMA handles clinical abbreviations\, recognizes implic it phenotype patterns\, and applies contextual reasoning locally on stand ard hardware to extract and code rare disease information with supporting textual evidence. This approach reduces privacy risks while improving F1 performance by over 30\% and decreasing inference costs 10-fold\, achiev ing high precision (89\%) in rare disease mining and coding. By enabling clinicians to access systematically coded rare disease information with e xplicit evidence from EHR systems without cloud-based privacy risks\, RDM A supports identification and documentation of rare conditions. Available at https://github.com/jhnwu3/RDMA. John Wu is a a Ph.D student in the CS department at the University of Illinois\, currently advised by Professo r Jimeng Sun . His main focus is on building agentic systems for healthca re settings\, whether that be low resource (i.e rare diseases)\, interpre tability\, or clinical predictive modeling. He actively maintains PyHealt h \, and leads a community of open-source researchers\, trying to build m ore reproducible healthcare solutions. His work is currently supported by the NSF GRFP.\n\nSpeaker:\nJohn Wu\n\nAdmission:\nFree\n\nDetails URL:\n https://medicine.yale.edu/event/nlpllm-interest-group-johnwu/\n DTEND;TZID=America/New_York:20251110T170000 DTSTAMP:20260429T090341Z DTSTART;TZID=America/New_York:20251110T160000 LOCATION:Join via Zoom: https://yale.zoom.us/j/91284647043?pwd=SHB0bu8FRja YFhJmpxYRKUDXlINIYA.1\n\nPassword: 159239\n\nOr Telephone：203-432-9666 (2 -ZOOM if on-campus) or 646 568 7788\n One Tap Mobile: +12034329666\,\, 91284647043# US (Bridgeport)\n\n Meeting ID: 912 8464 7043\n Intern ational numbers available: https://yale.zoom.us/u/ac8x2d0qLF\n\, URL: htt ps://yale.zoom.us/j/91284647043?pwd=SHB0bu8FRjaYFhJmpxYRKUDXlINIYA.1 RECURRENCE-ID;TZID=America/New_York:20251110T160000 SEQUENCE:0 STATUS:Confirmed SUMMARY:NLP/LLM Interest Group: RDMA UID:e776ef46-f30b-4de3-80b8-65f255e3ba90 END:VEVENT BEGIN:VEVENT DESCRIPTION:\nAdmission:\nFree\n\nDetails URL:\nhttps://medicine.yale.edu/ event/nlpllm-interest-group-10/\n DTEND;TZID=America/New_York:20251117T170000 DTSTAMP:20260429T090341Z DTSTART;TZID=America/New_York:20251117T160000 LOCATION:Zoom: Passcode Required\, URL: https://yale.zoom.us/j/93599941969 RECURRENCE-ID;TZID=America/New_York:20251117T160000 SEQUENCE:0 STATUS:Cancelled SUMMARY:NLP/LLM Interest Group UID:e776ef46-f30b-4de3-80b8-65f255e3ba90 END:VEVENT BEGIN:VEVENT DESCRIPTION:\nAdmission:\nFree\n\nDetails URL:\nhttps://medicine.yale.edu/ event/nlpllm-interest-group-11/\n DTEND;TZID=America/New_York:20251124T170000 DTSTAMP:20260429T090341Z DTSTART;TZID=America/New_York:20251124T160000 LOCATION:Zoom: Passcode Required\, URL: https://yale.zoom.us/j/93599941969 RECURRENCE-ID;TZID=America/New_York:20251124T160000 SEQUENCE:0 STATUS:Cancelled SUMMARY:NLP/LLM Interest Group UID:e776ef46-f30b-4de3-80b8-65f255e3ba90 END:VEVENT BEGIN:VEVENT DESCRIPTION:This session will feature two exciting talks: 1. Rethinking Re trieval-Augmented Generation for Medicine: A Large-Scale\, Systematic Exp ert Evaluation and Practical Insights by Hyunjae Kim\, PhD Abstract: Retr ieval-augmented generation (RAG) is widely adopted to keep medical LLMs c urrent and verifiable\, yet its effectiveness remains unclear. We present the first end-to-end\, expert annotated evaluation of RAG in medicine\, systematically assessing the full pipeline across three stages: evidence retrieval\, evidence selection\, and response generation. Eighteen medica l experts provided 80\,502 annotations across 800 model outputs on 200 cl inical queries.Contrary to expectations\, conventional RAG often degraded performance—only 22% of retrieved passages were relevant\, evidence sele ction was weak\, and factuality dropped up to 6%. However\, simple strate gies like evidence filtering and query reformulation improved performance by up to 12%. Our findings challenge current RAG assumptions and highlig ht the need for deliberate system design in medical AI applications. 2. T opicForest: Embedding-Driven Hierarchical Clustering and Labeling for Bio medical Literature by Chia-Hsuan Chang\, PhD Abstract: The vast and compl ex landscape of biomedical literature presents significant challenges for organization and interpretation. Current embedding-based topic models li ke BERTopic are limited to flat\, single-granularity clusters\, failing t o capture the inherently nested\, hierarchical structure of scientific su bjects. We introduce TopicForest\, a novel framework that captures this n atural hierarchy by building a "forest of topic trees" directly from text embeddings. TopicForest delivers high-quality topic clustering comparabl e to state-of-the-art flat models while providing the essential multi-sca le resolution they lack. Through recursive topic labeling\, the framework achieves efficient token usage and practical scalability for large corpo ra. This design provides researchers with an effective tool for exploring and visualizing hierarchical biomedical knowledge landscapes.\n\nSpeaker s:\nHyunjae Kim\; Chia-Hsuan Chang\n\nAdmission:\nFree\n\nDetails URL:\nh ttps://medicine.yale.edu/event/nlpllm-interest-group-12/\n DTEND;TZID=America/New_York:20251201T170000 DTSTAMP:20260429T090341Z DTSTART;TZID=America/New_York:20251201T160000 LOCATION:Zoom: Passcode Required\, URL: https://yale.zoom.us/j/93599941969 RECURRENCE-ID;TZID=America/New_York:20251201T160000 SEQUENCE:0 STATUS:Confirmed SUMMARY:NLP/LLM Interest Group UID:e776ef46-f30b-4de3-80b8-65f255e3ba90 END:VEVENT BEGIN:VEVENT DESCRIPTION:This session will feature two exciting talks: A Collaborative Reasoning Agent-based Framework with Built-in Verification for Safe Medic al Decision-Making by Fan Ma\, PhD Abstract: Large language models (LLMs) have demonstrated expert-level capabilities on medical benchmarks\, yet translating these achievements into clinical practice is impeded by persi stent risks of hallucination and a lack of verifiable reasoning. While em erging agentic frameworks have begun to address these limitations through multi-step planning\, existing systems often prioritize performance opti mization over rigorous safety checks and fail to emulate the collective d ecision-making of multidisciplinary teams. To address these critical gaps \, we introduce OpenDx\, a multi-agent framework designed to bridge the d ivide between experimental prototypes and reliable clinical decision supp ort. OpenDx is built upon three core principles: collaboration among spec ialized agents that simulate distinct clinical roles\, integrated verific ation modules that strictly cross-check outputs for safety and consistenc y\, and an architectural alignment with clinical auditability standards. We present the design and evaluation of OpenDx\, demonstrating how struct ured collaboration significantly enhance reliability compared to baseline models. Our work advocates for a new paradigm of trustworthy medical AI\ , where performance gains are inseparable from the interpretability and s afety assurances required for frontline healthcare deployment. A Federate d and Parameter-Efficient Framework for Large Language Model Training in Medicine: Applications to Clinical Information Extraction by Anran Li\, P hD Abstract: Large language models (LLMs) are advancing medical applicati ons such as patient question answering and diagnosis. Yet extracting stru ctured information from unstructured clinical narratives across healthcar e systems remains challenging. Current LLMs struggle with such clinical i nformation extraction (IE) due to complex language\, limited annotations\ , and data silos. We present a federated\, model-agnostic framework for t raining LLMs in medicine\, applied to clinical IE. The proposed Fed-MedLo RA enables parameter-efficient federated fine-tuning by transmitting only low-rank adapter parameters\, substantially reducing communication and c omputation costs. Accuracy was assessed across five patient cohorts throu gh comparisons with baselines for LLMs under (1) in-domain training and t esting\, (2) external patient cohorts\, and (3) a case study on new-site adaptation using real-world clinical notes.\n\nSpeaker:\nFan Ma\n\nAdmiss ion:\nFree\n\nDetails URL:\nhttps://medicine.yale.edu/event/nlpllm-intere st-group-13/\n DTEND;TZID=America/New_York:20251208T170000 DTSTAMP:20260429T090341Z DTSTART;TZID=America/New_York:20251208T160000 LOCATION:Zoom: Passcode Required\, URL: https://yale.zoom.us/j/93599941969 RECURRENCE-ID;TZID=America/New_York:20251208T160000 SEQUENCE:0 STATUS:Confirmed SUMMARY:NLP/LLM Interest Group UID:e776ef46-f30b-4de3-80b8-65f255e3ba90 END:VEVENT BEGIN:VEVENT DESCRIPTION:A Prompt Library for Efficient Clinical Entity Recognition Usi ng Large Language Models by Yang Ren\, PhD Abstract: Large Language Model s (LLMs) hold strong potential for clinical information extraction (IE)\, but their evaluation is often limited by manually crafted prompts and th e need for annotated data. We developed an automated framework that extra cts entity-level schema information from published clinical IE studies to construct structured prompts. Using literature covering 44 diseases and over 100 entities\, we generated prompts to evaluate multiple LLMs under few-shot and fine-tuned settings. Compared to baselines using generic pro mpts\, models prompted with schema-derived information consistently outpe rformed across tasks. Our results demonstrate the value of structured pro mpting for robust and reproducible LLM evaluation in diverse clinical IE applications.\n\nAdmission:\nFree\n\nDetails URL:\nhttps://medicine.yale. edu/event/nlpllm-interest-group-14/\n DTEND;TZID=America/New_York:20251215T170000 DTSTAMP:20260429T090341Z DTSTART;TZID=America/New_York:20251215T160000 LOCATION:Zoom: Passcode Required - Please Email Sooyoun Tan \, URL: https ://yale.zoom.us/j/93599941969 RECURRENCE-ID;TZID=America/New_York:20251215T160000 SEQUENCE:0 STATUS:Confirmed SUMMARY:NLP/LLM Interest Group UID:e776ef46-f30b-4de3-80b8-65f255e3ba90 END:VEVENT BEGIN:VEVENT DESCRIPTION:This session will feature two exciting talks: 1. Accelerating Cohort Identification from EHRs with Biomedical Knowledge and LLMs by Lin gfei Qian\, PhD Abstract ： Identifying eligible patients from electronic health records (EHRs) is a key challenge in clinical research. We present a framework that combines large language models (LLMs)\, Text-to-SQL\, a nd retrieval-augmented generation (RAG) to streamline cohort identificati on. Eligibility criteria are first decomposed and partially translated in to structured queries via Text-to-SQL\, providing a preliminary selection from OMOP-formatted EHR data. The core innovation focuses on RAG/QA to r etrieve and assess patient-level evidence from both clinical notes and st ructured tables\, emphasizing nuanced evaluation of complex criteria like disease chronicity\, lab thresholds\, and clinical stability\, while sup porting interactive cohort exploration and detailed patient-level evidenc e review. This workflow reduces manual effort\, improves accuracy\, and o ffers a scalable\, clinically grounded solution for EHR-based cohort iden tification. 2. An Information Extraction Approach to Detecting Novelty of Biomedical Publications by Xueqing Peng\, PhD Abstract : Scientific nove lty plays a critical role in shaping research impact\, yet it remains inc onsistently defined and difficult to quantify. Existing approaches often reduce novelty to a single measure\, failing to distinguish the specific types of contributions (such as new concepts or relationships) that drive influence. In this study\, we introduce a semantic measure of novelty ba sed on the emergence of new biomedical entities and relationships within the conclusion sections of research articles. Leveraging transformer-base d named entity recognition (NER) and relation extraction (RE) tools\, we identify novel findings and classify articles into four categories: No No velty\, Entity-only Novelty\, Relation-only Novelty\, and Entity-Relation Novelty. We evaluate this framework using citation counts and Journal Im pact Factors (JIF) as proxies for research influence. Our results show th at Entity-Relation Novelty articles receive the highest citation impact\, with relation novelty more closely aligned with high-impact journals. Th ese findings offer a scalable framework for assessing novelty and guiding future research evaluation.\n\nSpeakers:\nLingfei Qian\; Xueqing Peng\n\ nAdmission:\nFree\n\nDetails URL:\nhttps://medicine.yale.edu/event/nlpllm -interest-group-15/\n DTEND;TZID=America/New_York:20251222T170000 DTSTAMP:20260429T090341Z DTSTART;TZID=America/New_York:20251222T160000 LOCATION:Zoom: Passcode Required\, URL: https://yale.zoom.us/j/93599941969 RECURRENCE-ID;TZID=America/New_York:20251222T160000 SEQUENCE:0 STATUS:Confirmed SUMMARY:NLP/LLM Interest Group UID:e776ef46-f30b-4de3-80b8-65f255e3ba90 END:VEVENT BEGIN:VEVENT DESCRIPTION:\nAdmission:\nFree\n\nDetails URL:\nhttps://medicine.yale.edu/ event/nlpllm-interest-group-16/\n DTEND;TZID=America/New_York:20251229T170000 DTSTAMP:20260429T090341Z DTSTART;TZID=America/New_York:20251229T160000 LOCATION:Zoom: Passcode Required\, URL: https://yale.zoom.us/j/93599941969 RECURRENCE-ID;TZID=America/New_York:20251229T160000 SEQUENCE:0 STATUS:Cancelled SUMMARY:NLP/LLM Interest Group UID:e776ef46-f30b-4de3-80b8-65f255e3ba90 END:VEVENT BEGIN:VEVENT DESCRIPTION:\nAdmission:\nFree\n\nDetails URL:\nhttps://medicine.yale.edu/ event/nlpllm-interest-group-17/\n DTEND;TZID=America/New_York:20260105T170000 DTSTAMP:20260429T090341Z DTSTART;TZID=America/New_York:20260105T160000 LOCATION:Zoom: Passcode Required\, URL: https://yale.zoom.us/j/93599941969 RECURRENCE-ID;TZID=America/New_York:20260105T160000 SEQUENCE:0 STATUS:Confirmed SUMMARY:NLP/LLM Interest Group UID:e776ef46-f30b-4de3-80b8-65f255e3ba90 END:VEVENT BEGIN:VEVENT DESCRIPTION:\nAdmission:\nFree\n\nDetails URL:\nhttps://medicine.yale.edu/ event/nlpllm-interest-group-18/\n DTEND;TZID=America/New_York:20260112T170000 DTSTAMP:20260429T090341Z DTSTART;TZID=America/New_York:20260112T160000 LOCATION:Zoom: Passcode Required\, URL: https://yale.zoom.us/j/93599941969 RECURRENCE-ID;TZID=America/New_York:20260112T160000 SEQUENCE:0 STATUS:Cancelled SUMMARY:NLP/LLM Interest Group UID:e776ef46-f30b-4de3-80b8-65f255e3ba90 END:VEVENT BEGIN:VEVENT DESCRIPTION:\nAdmission:\nFree\n\nDetails URL:\nhttps://medicine.yale.edu/ event/nlpllm-interest-group-19/\n DTEND;TZID=America/New_York:20260119T170000 DTSTAMP:20260429T090341Z DTSTART;TZID=America/New_York:20260119T160000 LOCATION:Zoom: Passcode Required\, URL: https://yale.zoom.us/j/93599941969 RECURRENCE-ID;TZID=America/New_York:20260119T160000 SEQUENCE:0 STATUS:Tentative SUMMARY:NLP/LLM Interest Group UID:e776ef46-f30b-4de3-80b8-65f255e3ba90 END:VEVENT BEGIN:VEVENT DESCRIPTION:\nAdmission:\nFree\n\nDetails URL:\nhttps://medicine.yale.edu/ event/nlpllm-interest-group-20/\n DTEND;TZID=America/New_York:20260126T170000 DTSTAMP:20260429T090341Z DTSTART;TZID=America/New_York:20260126T160000 LOCATION:Zoom: Passcode Required\, URL: https://yale.zoom.us/j/93599941969 RECURRENCE-ID;TZID=America/New_York:20260126T160000 SEQUENCE:0 STATUS:Cancelled SUMMARY:NLP/LLM Interest Group UID:e776ef46-f30b-4de3-80b8-65f255e3ba90 END:VEVENT BEGIN:VEVENT DESCRIPTION:\nAdmission:\nFree\n\nDetails URL:\nhttps://medicine.yale.edu/ event/nlpllm-interest-group-21/\n DTEND;TZID=America/New_York:20260202T170000 DTSTAMP:20260429T090341Z DTSTART;TZID=America/New_York:20260202T160000 LOCATION:Zoom: Passcode Required\, URL: https://yale.zoom.us/j/93599941969 RECURRENCE-ID;TZID=America/New_York:20260202T160000 SEQUENCE:0 STATUS:Cancelled SUMMARY:NLP/LLM Interest Group UID:e776ef46-f30b-4de3-80b8-65f255e3ba90 END:VEVENT BEGIN:VEVENT DESCRIPTION:Title: Diagnostic Accuracy and Clinical Reasoning of Multiple Large Language Models Abstract: Large language models are increasingly us ed for mental health–related questions\, yet their performance in psychia try - where diagnosis depends heavily on narrative interpretation and cli nical reasoning - remains poorly understood. In this talk\, I’ll present a mixed-methods evaluation of four contemporary LLMs on 196 psychiatric c ase vignettes\, combining large-scale diagnostic accuracy metrics with cl inician-rated assessments of diagnostic reasoning. We find that models ca n achieve high diagnostic accuracy on vignettes\, but - crucially - that clinician-rated reasoning quality is far more predictive of diagnostic co rrectness than surface-level data extraction. These findings suggest that evaluating how models reason\, not just what they predict\, is essential for understanding their potential role in psychiatric decision support. Kevin Jin is a third-year PhD student in the Interdepartmental Program in Computational Biology and Biomedical Informatics at Yale University. He is advised by Hua Xu in the Clinical NLP Lab\, a research group in the De partment of Biomedical Informatics and Data Science at Yale School of Med icine. He completed his undergraduate work at Johns Hopkins University\, receiving a B.S. in Molecular and Cellular Biology in 2020. He is support ed by the NSF Graduate Research Fellowship.\n\nSpeaker:\nKevin Jin\n\nAdm ission:\nFree\n\nDetails URL:\nhttps://medicine.yale.edu/event/nlpllm-int erest-group-22/\n DTEND;TZID=America/New_York:20260209T170000 DTSTAMP:20260429T090341Z DTSTART;TZID=America/New_York:20260209T160000 LOCATION:Zoom link and passcode will be share on an email RECURRENCE-ID;TZID=America/New_York:20260209T160000 SEQUENCE:0 STATUS:Confirmed SUMMARY:NLP/LLM Interest Group UID:e776ef46-f30b-4de3-80b8-65f255e3ba90 END:VEVENT BEGIN:VEVENT DESCRIPTION:Title: Rethinking User Interface Design in the Era of AI Agent s Abstract: Artificial intelligence agents are rapidly reshaping how user s interact with digital systems. From embedded copilots to autonomous tas k executors\, agents are no longer confined to chat interfaces—they are b ecoming integral components of modern user interfaces. In this talk\, we will share a series of real-world cases and practical lessons drawn from building agent-driven systems in research and data-intensive environments . We will examine how agents are currently embedded into interfaces\, wha t architectural decisions influence usability and trust\, and what design trade-offs emerge when combining autonomy with human control. We will al so discuss how AI agent tools themselves are transforming the UI design w orkflow—from rapid prototyping to code generation and interaction simulat ion. Huan He\, PhD \, is a research scientist in biomedical informatics a nd data science at Yale University School of Medicine. His primary resear ch areas revolve around visual analytics and their applications in health care-related research. Currently\, his work is focused on designing and d eveloping visual analytics systems using natural language processing (NLP ) and machine learning (ML) technologies\, with the goal of facilitating data exploration for health-related clinical questions.\n\nSpeaker:\nHuan He\n\nAdmission:\nFree\n\nDetails URL:\nhttps://medicine.yale.edu/event/ nlpllm-interest-group-23/\n DTEND;TZID=America/New_York:20260216T170000 DTSTAMP:20260429T090341Z DTSTART;TZID=America/New_York:20260216T160000 LOCATION:Zoom: Passcode Required\, URL: https://yale.zoom.us/j/93599941969 RECURRENCE-ID;TZID=America/New_York:20260216T160000 SEQUENCE:0 STATUS:Confirmed SUMMARY:NLP/LLM Interest Group UID:e776ef46-f30b-4de3-80b8-65f255e3ba90 END:VEVENT BEGIN:VEVENT DESCRIPTION:Title: HypRAG: Hyperbolic Dense Retrieval for Retrieval Augmen ted Generation Abstract: Embedding geometry fundamentally affects retriev al quality\, yet dense retrievers for retrieval-augmented generation (RAG ) remain confined to Euclidean space. Natural language has hierarchical s tructure from broad topics to specific entities that Euclidean embeddings fail to preserve\, causing semantically distant documents to appear spur iously similar and increasing hallucination risk. We introduce hyperbolic dense retrieval with two model variants in the Lorentz model: HyTE-FH\, a fully hyperbolic transformer\, and HyTE-H\, a hybrid architecture proje cting pre-trained Euclidean embeddings into hyperbolic space. To prevent representational collapse during sequence aggregation\, we introduce the Outward Einstein Midpoint\, a geometry-aware pooling operator that provab ly preserves hierarchical structure. On MTEB\, HyTE-FH outperforms equiva lent Euclidean baselines. On RAGBench\, HyTE-H achieves up to 29% gains o ver Euclidean baselines in context and answer relevance using substantial ly smaller models than current state-of-the-art retrievers. Hyperbolic re presentations encode document specificity through norm-based separation\, with over 20% radial increase from general to specific concepts\, a prop erty absent in Euclidean embeddings. Hiren Madhu a second year PhD studen t at Yale CS \, advised by Professor Smita Krishnaswamy and Professor Rex Ying . Earlier\, he was an ECE pre-doctoral research fellow in Electrica l Communication Engineering Department at the Indian Institute of Science \, Bengaluru. He worked with Sundeep Prabhakar Chepuri on unsupervised re presentation learning for simplicial complexes. He began his research jou rney at LDRP Institute of Technology and Research (BE in Computer Enginee ring) under the guidance of Sandip Modha and Thomas Mandl\, where he focu sed on the detection of hate speech in public conversational threads on s ocial media. His primary research interest is encoding the geometric indu ctive biases into foundation models. Previously\, he worked on simplicial representation learning without labels.\n\nSpeaker:\nHiren Madhu\n\nAdmi ssion:\nFree\n\nDetails URL:\nhttps://medicine.yale.edu/event/nlpllm-inte rest-group-25/\n DTEND;TZID=America/New_York:20260302T170000 DTSTAMP:20260429T090341Z DTSTART;TZID=America/New_York:20260302T160000 LOCATION:Zoom: Passcode Required\, URL: https://yale.zoom.us/j/93599941969 RECURRENCE-ID;TZID=America/New_York:20260302T160000 SEQUENCE:0 STATUS:Confirmed SUMMARY:NLP/LLM Interest Group UID:e776ef46-f30b-4de3-80b8-65f255e3ba90 END:VEVENT BEGIN:VEVENT DESCRIPTION:Title: OpenClaw: A Prospective View on its Current Abilities a nd Potential Medical Field Application Abstract: OpenClaw (formerly MoltB ot) is one of the newest AI technologies to grab the attention of develop ers\, business owners\, and tech enthusiasts across the internet. Perhaps most profoundly recognized for supporting the bots behind MoltBook---the social media platform entirely populated and used by AI---OpenClaw is ad vertised as the next step in AI’s developmental progression\, where model s stop being only chatbots and gain the capability to undertake delegated tasks fully from start to finish. As an autonomous virtual AI assistant\ , OpenClaw brings many potential use cases to the medical field\, along w ith certain concerns of its security and potential for malicious activity . This discussion will explore considerable applications to both clinical scenarios and tech-infrastructure support. Jeffrey Li is an alumnus of B oston University\, graduating in 2025 with majors in Computer Science and Economics.\n\nAdmission:\nFree\n\nDetails URL:\nhttps://medicine.yale.ed u/event/nlpllm-interest-group-26/\n DTEND;TZID=America/New_York:20260309T170000 DTSTAMP:20260429T090341Z DTSTART;TZID=America/New_York:20260309T160000 LOCATION:Zoom: Passcode Required\, URL: https://yale.zoom.us/j/93599941969 RECURRENCE-ID;TZID=America/New_York:20260309T160000 SEQUENCE:0 STATUS:Confirmed SUMMARY:NLP/LLM Interest Group UID:e776ef46-f30b-4de3-80b8-65f255e3ba90 END:VEVENT BEGIN:VEVENT DESCRIPTION:\nAdmission:\nFree\n\nDetails URL:\nhttps://medicine.yale.edu/ event/nlpllm-interest-group-27/\n DTEND;TZID=America/New_York:20260316T170000 DTSTAMP:20260429T090341Z DTSTART;TZID=America/New_York:20260316T160000 LOCATION:Zoom: Passcode Required\, URL: https://yale.zoom.us/j/93599941969 RECURRENCE-ID;TZID=America/New_York:20260316T160000 SEQUENCE:0 STATUS:Cancelled SUMMARY:NLP/LLM Interest Group UID:e776ef46-f30b-4de3-80b8-65f255e3ba90 END:VEVENT BEGIN:VEVENT DESCRIPTION:This session features two talks exploring novel approaches to AI-driven biomedical discovery and drug representation. "PredMed: Extrapo lating Future Discoveries from the Literature Universe" by Chia-Hsuan Cha ng\, PhD - Postdoctoral Associate in Biomedical Informatics and Data Scie nce Abstract: Most AI co-scientists are limited by retrieval-augmented ge neration (RAG) over static corpora and heavily rely on human guidance. We present PredMed\, a novel framework that redefines hypothesis generation as a temporal extrapolation task within the high-dimensional literature universe. Using time-based regression and a specialized Embedding Languag e Model (ELM) acting as a decoder\, we project and translate future-state embeddings back into natural language. Our results show that this tempor al steering mechanism explores scientific territory that standard prompti ng cannot reach\, outperforming baseline methods in both novelty and rela tional depth. We also validate PredMed’s efficacy through expert-reviewed hypotheses in CAR-T therapy domain\, highlighting a new frontier for aut onomous scientific discovery "A Literature-Based Drug Embedding Resource for Biomedical Research" by Zhiyuan Cao - PhD Student in Computational Bi ology and Biomedical Informatics (CBB) Abstract: We introduce DrugSpace\, a reusable text-based drug embedding resource designed to support simila rity search\, retrieval\, and downstream modeling in biomedical research. Built from large-scale PubMed abstracts and aligned with heterogeneous D rugBank drug descriptions through a two-stage training pipeline\, DrugSpa ce is released both as a versioned embedding dataset and as an embedder f or generating representations from new drug text. To support realistic re use\, the resource is evaluated under a prospective setting that separate s drug-level alignment from later drug introductions and updates. Across intrinsic similarity discrimination\, ATC-based therapeutic retrieval\, r obustness to input perturbations\, and integration into a representative DDI prediction pipeline\, DrugSpace consistently remains competitive with strong biomedical and general-purpose text embedding baselines\, support ing its utility as a practical and extensible drug representation resourc e.\n\nSpeakers:\nChia-Hsuan Chang\; Zhiyuan Cao\n\nAdmission:\nFree\n\nDe tails URL:\nhttps://medicine.yale.edu/event/nlpllm-interest-group-28/\n DTEND;TZID=America/New_York:20260323T170000 DTSTAMP:20260429T090341Z DTSTART;TZID=America/New_York:20260323T160000 LOCATION:Subscribe to receive Zoom Passcode: https://mailman.yale.edu/mail man/listinfo/nlp-llm-ig \, URL: https://yale.zoom.us/j/93599941969 RECURRENCE-ID;TZID=America/New_York:20260323T160000 SEQUENCE:0 STATUS:Confirmed SUMMARY:NLP/LLM Interest Group UID:e776ef46-f30b-4de3-80b8-65f255e3ba90 END:VEVENT BEGIN:VEVENT DESCRIPTION:This session features two talks: "A Systematic Framework for S caling High-Fidelity Medical Multimodal Data Curation" by Hyunjae Kim\, P hD - Postdoctoral Associate in Biomedical Informatics and Data Science Ab stract: Medical multimodal learning is often hindered by the scarcity of high-quality image-text data. While scientific literature is a vast resou rce\, extracting clinically relevant\, deconstructed\, and aligned image- caption pairs at scale remains challenging. We propose MedPMC\, an automa ted five-stage pipeline to curate high-fidelity medical datasets from lar ge-scale biomedical repositories. The framework features task-specialized components for precise image filtering and sophisticated separation of m ulti-panel figures and captions to ensure accurate image-text alignment. Leveraging this pipeline\, we curated 12 million medical image-text pairs . A CLIP-style model trained on this dataset surpassed state-of-the-art p erformance across 20+ benchmarks in six clinical specialties\, including radiology and pathology. Furthermore\, integrating our model into a multi modal LLM outperformed baselines on medical QA tasks by 3.6%. Crucially\, MedPMC-trained models enhance performance on internal clinical data\, un derscoring their utility in real-world settings. This scalable framework establishes a new paradigm for transforming biomedical literature into co ntinuously updatable\, clinically grounded training resources. " S-index – A Refined Data Sharing Index to Promote and Reward Biomedical Data Reus e" by Kalpana Raja\, PhD\, MRSB\, CSci - Instructor of Biomedical Informa tics and Data Science Abstract: Data sharing has become increasingly reco gnized as essential for accelerating scientific discovery\, enhancing tra nsparency\, and maximizing the return on research investments. Efforts su ch as the FAIR principles (Findable\, Accessible\, Interoperable\, and Re usable) and NIH policies on Data Management and Sharing have underscored the importance of making datasets widely available to the scientific comm unity. Despite these advances\, current practices lack quantitative metri cs that accurately reflect researchers' contributions to data sharing\, p articularly the downstream reuse of their datasets by the broader scienti fic community. Existing citation metrics\, such as the H-index\, predomin antly measure scholarly impact through publications\, neglecting the crit ical role of dataset creation and reuse. Consequently\, there is a pressi ng need for a novel index to quantify and incentivize dataset reuse\, fos tering a robust culture of open and impactful scientific data sharing. To address this\, we propose the Data Sharing index (S-index)\, a refined m etric specifically designed to quantify a researcher’s contribution to re usable data. We have built an end-to-end workflow for S-index computation and a web-based interface for visualization and demonstrated feasibility using a real-world repository (OpenNeuro).\n\nSpeakers:\nHyunjae Kim\; K alpana Raja\n\nAdmission:\nFree\n\nDetails URL:\nhttps://medicine.yale.ed u/event/nlpllm-interest-group-29/\n DTEND;TZID=America/New_York:20260330T170000 DTSTAMP:20260429T090341Z DTSTART;TZID=America/New_York:20260330T160000 LOCATION:Join our mailing list to receive Zoom Passcode: https://mailman.y ale.edu/mailman/listinfo/nlp-llm-ig\, URL: https://yale.zoom.us/j/9359994 1969 RECURRENCE-ID;TZID=America/New_York:20260330T160000 SEQUENCE:0 STATUS:Confirmed SUMMARY:NLP/LLM Interest Group UID:e776ef46-f30b-4de3-80b8-65f255e3ba90 END:VEVENT BEGIN:VEVENT DESCRIPTION:Large language models have achieved near-saturated performance on medical knowledge benchmarks\, yet high exam scores tell us little ab out clinical safety or real-world utility. In this talk\, I review three recent studies that collectively reframe how we should evaluate LLMs for clinical use: NOHARM\, which introduces a safety-oriented benchmark revea ling that most LLM errors are harmful omissions rather than commissions\; MedR-Bench\, which decomposes clinical reasoning into stages and exposes critical weaknesses beyond diagnosis\; and the first randomized controll ed trial of ambient AI scribes\, which highlights the gap between technic al capability and clinical adoption. Together\, these works suggest a par adigm shift\, from asking "are LLMs smart enough for medicine?" to "how d o we rigorously evaluate their safety\, understand their failure modes\, and validate their real-world impact".\n\nSpeaker:\nLingfei Qian\n\nAdmis sion:\nFree\n\nDetails URL:\nhttps://medicine.yale.edu/event/nlpllm-inter est-group-30/\n DTEND;TZID=America/New_York:20260406T170000 DTSTAMP:20260429T090341Z DTSTART;TZID=America/New_York:20260406T160000 LOCATION:Join our mailing list to receive Zoom Passcode: https://mailman.y ale.edu/mailman/listinfo/nlp-llm-ig\, URL: https://yale.zoom.us/j/9359994 1969 RECURRENCE-ID;TZID=America/New_York:20260406T160000 SEQUENCE:0 STATUS:Confirmed SUMMARY:NLP/LLM Interest Group UID:e776ef46-f30b-4de3-80b8-65f255e3ba90 END:VEVENT BEGIN:VEVENT DESCRIPTION:In this presentation\, Michael Krauthammer\, MD\, PhD \, inves tigates the transition of Large Language Models (LLMs) and their underlyi ng technologies into clinical reasoning systems\, drawing on research and implementation examples from the University of Zurich and University Hos pital of Zurich. Dr. Krauthammer will first demonstrate the efficacy of Vis ion Transformers (ViT) and Vision-Language Models (VLM) for image assessm ent and reporting in rheumatology and radiology. He will then explore the use of LLMs in the Zurich AI tumor board for clinical guideline mapping and counterfactual treatment planning. Finally\, he will address the key hurdles to clinical adoption\, including stakeholder management\, infrast ructure requirements\, and certification.\n\nSpeaker:\nMichael Krauthamme r\, MD\, PhD\n\nAdmission:\nFree\n\nDetails URL:\nhttps://medicine.yale.e du/event/nlpllm-interest-group-31/\n DTEND;TZID=America/New_York:20260413T170000 DTSTAMP:20260429T090341Z DTSTART;TZID=America/New_York:20260413T160000 LOCATION:Join our mailing list to receive Zoom Passcode: https://mailman.y ale.edu/mailman/listinfo/nlp-llm-ig\, URL: https://yale.zoom.us/j/9359994 1969 RECURRENCE-ID;TZID=America/New_York:20260413T160000 SEQUENCE:0 STATUS:Confirmed SUMMARY:NLP/LLM Interest Group UID:e776ef46-f30b-4de3-80b8-65f255e3ba90 END:VEVENT BEGIN:VEVENT DESCRIPTION:\nAdmission:\nFree\n\nDetails URL:\nhttps://medicine.yale.edu/ event/nlpllm-interest-group-32/\n DTEND;TZID=America/New_York:20260420T170000 DTSTAMP:20260429T090341Z DTSTART;TZID=America/New_York:20260420T160000 LOCATION:Zoom: Passcode Required\, URL: https://yale.zoom.us/j/93599941969 RECURRENCE-ID;TZID=America/New_York:20260420T160000 SEQUENCE:0 STATUS:Cancelled SUMMARY:NLP/LLM Interest Group UID:e776ef46-f30b-4de3-80b8-65f255e3ba90 END:VEVENT BEGIN:VEVENT DESCRIPTION:\nAdmission:\nFree\n\nDetails URL:\nhttps://medicine.yale.edu/ event/nlpllm-interest-group-33/\n DTEND;TZID=America/New_York:20260427T170000 DTSTAMP:20260429T090341Z DTSTART;TZID=America/New_York:20260427T160000 LOCATION:Zoom: Passcode Required\, URL: https://yale.zoom.us/j/93599941969 RECURRENCE-ID;TZID=America/New_York:20260427T160000 SEQUENCE:0 STATUS:Cancelled SUMMARY:NLP/LLM Interest Group UID:e776ef46-f30b-4de3-80b8-65f255e3ba90 END:VEVENT BEGIN:VEVENT DESCRIPTION:\nAdmission:\nFree\n\nDetails URL:\nhttps://medicine.yale.edu/ event/nlpllm-interest-group-34/\n DTEND;TZID=America/New_York:20260504T170000 DTSTAMP:20260429T090341Z DTSTART;TZID=America/New_York:20260504T160000 LOCATION:Zoom: Passcode Required\, URL: https://yale.zoom.us/j/93599941969 RECURRENCE-ID;TZID=America/New_York:20260504T160000 SEQUENCE:0 STATUS:Tentative SUMMARY:NLP/LLM Interest Group UID:e776ef46-f30b-4de3-80b8-65f255e3ba90 END:VEVENT BEGIN:VEVENT DESCRIPTION:\nAdmission:\nFree\n\nDetails URL:\nhttps://medicine.yale.edu/ event/nlpllm-interest-group-35/\n DTEND;TZID=America/New_York:20260511T170000 DTSTAMP:20260429T090341Z DTSTART;TZID=America/New_York:20260511T160000 LOCATION:Zoom: Passcode Required\, URL: https://yale.zoom.us/j/93599941969 RECURRENCE-ID;TZID=America/New_York:20260511T160000 SEQUENCE:0 STATUS:Tentative SUMMARY:NLP/LLM Interest Group UID:e776ef46-f30b-4de3-80b8-65f255e3ba90 END:VEVENT BEGIN:VEVENT DESCRIPTION:\nAdmission:\nFree\n\nDetails URL:\nhttps://medicine.yale.edu/ event/nlpllm-interest-group-36/\n DTEND;TZID=America/New_York:20260518T170000 DTSTAMP:20260429T090341Z DTSTART;TZID=America/New_York:20260518T160000 LOCATION:Zoom: Passcode Required\, URL: https://yale.zoom.us/j/93599941969 RECURRENCE-ID;TZID=America/New_York:20260518T160000 SEQUENCE:0 STATUS:Tentative SUMMARY:NLP/LLM Interest Group UID:e776ef46-f30b-4de3-80b8-65f255e3ba90 END:VEVENT BEGIN:VEVENT DESCRIPTION:\nAdmission:\nFree\n\nDetails URL:\nhttps://medicine.yale.edu/ event/nlpllm-interest-group-37/\n DTEND;TZID=America/New_York:20260525T170000 DTSTAMP:20260429T090341Z DTSTART;TZID=America/New_York:20260525T160000 LOCATION:Zoom: Passcode Required\, URL: https://yale.zoom.us/j/93599941969 RECURRENCE-ID;TZID=America/New_York:20260525T160000 SEQUENCE:0 STATUS:Tentative SUMMARY:NLP/LLM Interest Group UID:e776ef46-f30b-4de3-80b8-65f255e3ba90 END:VEVENT BEGIN:VEVENT DESCRIPTION:\nAdmission:\nFree\n\nDetails URL:\nhttps://medicine.yale.edu/ event/nlpllm-interest-group-2/\n DTEND;TZID=America/New_York:20260223T170000 DTSTAMP:20260429T090341Z DTSTART;TZID=America/New_York:20260223T160000 LOCATION:Zoom: Passcode Required\, URL: https://yale.zoom.us/j/93599941969 RECURRENCE-ID;TZID=America/New_York:20260223T160000 SEQUENCE:0 STATUS:Confirmed SUMMARY:NLP/LLM Interest Group UID:e776ef46-f30b-4de3-80b8-65f255e3ba90 END:VEVENT BEGIN:VEVENT DESCRIPTION:\nAdmission:\nFree\n DTEND;TZID=America/New_York:20250915T170000 DTSTAMP:20260429T090341Z DTSTART;TZID=America/New_York:20250915T160000 EXDATE:20260223T160000 LOCATION:Zoom: Passcode Required\, URL: https://yale.zoom.us/j/93599941969 RRULE:FREQ=WEEKLY;UNTIL=20260526T035959Z;BYDAY=MO SEQUENCE:0 STATUS:Confirmed SUMMARY:NLP/LLM Interest Group UID:e776ef46-f30b-4de3-80b8-65f255e3ba90 END:VEVENT END:VCALENDAR