Also on Google Scholar. ∗ denotes equal contribution and † denotes the corresponding author.
GroundLM 2026 (EMNLP 2026 Workshop)
This work studies how pruning, quantization, and KV-cache compression interact, and combines them to run a 70B model with 10k-token prompts on a single A40.
Experience Retrieval-Augmentation with Electronic Health Records Enables Accurate Discharge QA
ACL 2026 Main
ExpRAG retrieves similar patient experiences from electronic health records to improve discharge-related clinical question answering.
Patient-Similarity Cohort Reasoning in Clinical Text-to-SQL
EACL 2026 Main (Oral)
ClinSQL evaluates models on clinically grounded text-to-SQL tasks that require temporal reasoning and patient-similarity cohort construction.
SciVer: Evaluating Foundation Models for Multimodal Scientific Claim Verification
ACL 2025 Main
SciVer benchmarks foundation models on 3,000 expert-annotated claims grounded in scientific text, charts, and tables.