I am a senior undergraduate at National Chengchi University (NCCU),
double majoring in Computer Science and Management Information Systems.
My research focuses on trustworthy foundation models (LLMs and VLMs) for high-stakes applications such as
healthcare and autonomous driving, spanning hallucination, interpretability, and
verifiable model repair.
I currently work with Prof. Kuan-Hao Huang at Texas A&M University
on hallucination in vision-language models, and collaborate with
Prof. Liyue Shen's group at the University of Michigan
on active clinical diagnostic agents. I am also a research assistant at
Far Eastern Memorial Hospital,
where my system MedPlan (ACL 2025) has been deployed. At NCCU, I work with Prof. Fang Yu on verifiable repair of
language models and with Prof. Shi-Sheng Sun on explainable trajectory prediction.
I am applying to CS PhD programs for Fall 2027.
News
Oct 2026Released NAME-Bench (paper, data): not every object mention is a hallucination claim. Under review at ARR.
Oct 2026Released FORGE (paper, code), verification-gated behavioral repair for LLMs. Under review at ICSE 2027.
Jul 2026MedAction (paper, code/data) accepted to COLM 2026.
Hallucination benchmarks treat every object a caption names as a claim that the object is present.
But in “a bench for people to sit on”, people is mentioned, not asserted.
I built NAME-Bench, 25,497 object mentions from five VLMs, labeled as concrete, hallucinated, or
non-assertive (depictive, purpose, negation, hedged). Every detector we tested flags non-assertive mentions as
hallucinations (14–80%), and correcting this is far harder on newer VLMs than on LLaVA-1.5.
MedAction: Active Multi-turn Clinical Diagnostic LLMs
Real diagnosis is a sequence of decisions: what to ask, which test to order, and when to stop.
I built a tree-structured distillation pipeline with two knowledge-graph-grounded consistency metrics to create
MedAction-32K (32,681 multi-turn diagnostic trajectories from 2,896 PMC cases). Our fine-tuned 8B model
is state of the art among open-source models on MedRBench. I am now extending it with reinforcement learning into an
agent that orders tests, interprets results, and decides when to stop.
MedPlan: Personalized Medical Plan Generation in Clinical Practice
A two-stage retrieval-augmented system that mirrors how clinicians write SOAP notes: it first generates an assessment,
then a patient-specific treatment plan, grounded in the hospital's real-world EHR records. The system is deployed at
the hospital and validated by attending physicians on real patient records (90%+ ICD-10 prediction accuracy), and led
to a Taiwan patent, the National Innovation Award,
and MOUs with clinics and HIS vendors. We are now integrating it with HL7 FHIR for interoperability with hospital
information systems.
WARP & FORGE: Verifiable Repair of Language Models
Fine-tuning can fix a model's mistakes but offers no guarantee about what else it breaks.
WARP repairs inner layers of Transformer models by solving a convex quadratic program, with provable
guarantees on the repaired samples, preservation of correct behavior, and certified robustness.
FORGE extends this to generative LLMs, reducing bias and toxicity with minimal perplexity cost, and
stays stable in low-resource settings where fine-tuning fails.
MedAction: Towards Active Multi-turn Clinical Diagnostic LLMsHsin-Ling Hsu, Zizheng Wang*, Donghua Zhang*, …, Chenwei Wu†, Liyue Shen†Third Conference on Language Modeling (COLM 2026), San Francisco, USA, 2026. ~29% acceptance rate.[paper][code/data]
MedPlan: A Two-Stage RAG-Based System for Personalized Medical Plan GenerationHsin-Ling Hsu*, Cong-Tinh Dao*, Luning Wang, Zitao Shuai, …, Fang-Ming Hung, Chenwei WuProceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (ACL 2025): Industry Track, Vienna, Austria, 2025. ~25.6% acceptance rate. 30+ citations.[paper]
RGPO: Ranking-Guided Preference Optimization for Reliable Clinical ReasoningChia-Hsuan Hsu*, Jun-En Ding*, Hsin-Ling Hsu, …, Fang-Ming HungIEEE Transactions on Artificial Intelligence (IEEE TAI, Q1 Journal), 2026.[paper]
KAP: MLLM-assisted OCR Text Enhancement for Hybrid Retrieval in Chinese Non-Narrative DocumentsHsin-Ling Hsu, Ping-Sheng Lin, Jing-Di Lin, Jengnan TzengProceedings of the FinTech in AI CUP Special Session, 18th NTCIR Conference, Tokyo, Japan, 2025.[paper][code]Oral presentation; travel grant from Taiwan's Ministry of Education ($2,500 USD); National Excellence Prize, AI CUP 2024.
Manuscripts
Not Every Mention Is a Claim: Rethinking Object Hallucination Detection in Vision-Language ModelsHsin-Ling Hsu*, Neeraja Kirtane*, Kuan-Hao HuangUnder review at ARR October 2026 (targeting NAACL 2027).[paper][data]
FORGE: Verification-Gated Behavioral Repair for Generative Language ModelsHsin-Ling Hsu, Min-Yu Chen, Nai-Chia Chen, Yan-Ru Chen, Yi-Ling Chang, Fang YuUnder review at ICSE 2027.[paper][code]
Reasoning Like an LLM, Running Like an LSTM: Concept Graph Distillation for Explainable Vehicle Trajectory PredictionHsin-Ling Hsu, …, Shi-Sheng SunIn preparation.
Patent
Electronic Device for Generating Personalized Assessment Content and Treatment PlanHsin-Ling Hsu, Jun-En Ding, Fang-Ming HungTaiwan Utility Model Patent No. M682510, granted May 2026. Assignee: Far Eastern Memorial Hospital.[patent]
AI Engineer (Part-Time) · LLM/Whisper R&D (transcription, inference acceleration, LoRA tuning); RAG pipelines and an LLM agent for addiction counseling.
Service & Honors
Reviewer: KDD 2026 AI4Sci Track, IEEE Transactions on Consumer Electronics, NeurIPS Workshop on GenAI for Health (2025, 2026), NeurIPS 2025 Workshop on Efficient Reasoning, Texas NLP Symposium 2026, WMW 2026.
