mlsanityData to 7 Oct 2026

Coverage-Controlled Preference Mining from Noisy Claim Verification for Evidence-Grounded Generation

arXiv 2603.10494 ↗First posted 11 Mar 2026cs.CL cs.LGcited by 0

On the map: LLM alignment / RLHF / preference optimisation → Medical LLM alignment

Abstract (arXiv)

Evidence-grounded generation produces summaries whose claims should be supported by supplied evidence, but claim-level verifiers provide noisy feedback and can reward models that simply say less. We study this problem in clinical Brief Hospital Course summarization, where outputs must remain grounded in patient-specific EHR evidence. We introduce VERI-DPO, a preference-mining framework that converts noisy claim verification into coverage-controlled summary-level preferences. For each evidence-window prompt, VERI-DPO samples multiple candidate summaries, decomposes them into claims, verifies each claim against patient evidence, and forms a preference pair only when the chosen summary has better aggregate verifier-estimated support while retaining comparable verifiable content. Standard Direct Preference Optimization then distills these pairs into a single-sample policy, avoiding inference-time reranking. On patient-disjoint MIMIC-III-Ext-VeriFact-BHC test data, VERI-DPO reduces Not Supported rates from 10.7% to 1.9% under the mining verifier and from 11.6% to 6.4% under a separately prompted GPT-4o judge. In 100 blinded pairwise assessments by two domain researchers, VERI-DPO is preferred over the base model 56 times versus 18 times for factual faithfulness. In a locked zero-shot MIMIC-IV-Ext-BHC transfer test with 1,000 patients and no model adaptation, VERI-DPO lowers Not Supported rates with nearly unchanged scored-claim counts. Multi-seed ablations show that verifier-guided pair construction drives the gains, while coverage and anti-degeneration controls prevent apparent factuality improvements from coming from shorter or less checkable outputs.

Key facts. No citing papers in the map yet (data through 7 Oct 2026). Placed in Medical LLM alignment, one of 10 directions in LLM alignment / RLHF / preference optimisation. Builds on 18 of its 29 references that are in the map. First posted on arXiv 11 Mar 2026.

Builds on

18 of 29 references.

References · 18

  1. 2022Training language models to follow instructions with human feedbackcited by 15,712
  2. 2024The Llama 3 Herd of Modelscited by 14,192
  3. 2020Retrieval-Augmented Generation for Knowledge-Intensive NLP Taskscited by 9,160
  4. 2023Judging LLM-as-a-Judge with MT-Bench and Chatbot Arenacited by 8,393
  5. 2023Direct Preference Optimization: Your Language Model is Secretly a Reward Modelcited by 7,916
  6. 2024GPT-4o System Cardcited by 4,467
  7. 2023QLoRA: Efficient Finetuning of Quantized LLMscited by 3,355
  8. 2019The Curious Case of Neural Text Degenerationcited by 2,881
  9. 2020Learning to summarize from human feedbackcited by 2,562
  10. 2018FEVER: a large-scale dataset for Fact Extraction and VERificationcited by 1,362
  11. 2020On Faithfulness and Factuality in Abstractive Summarizationcited by 895
  12. 2023RLAIF vs. RLHF: Scaling Reinforcement Learning from Human Feedback with AI Feedbackcited by 615
6 more
  1. 2019Evaluating the Factual Consistency of Abstractive Text Summarizationcited by 578
  2. 2020Fact or Fiction: Verifying Scientific Claimscited by 546
  3. 2023Chain-of-Verification Reduces Hallucination in Large Language Modelscited by 317
  4. 2022RARR: Researching and Revising What Language Models Say, Using Language Modelscited by 286
  5. 2024Med42-v2: A Suite of Clinical LLMscited by 89
  6. 2024A dataset and benchmark for hospital course summarization with adapted large language modelscited by 10

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2026EviGen: Predictive Evidence Scaffolding for Verifiable Clinical Rationale Generation
  2. 2026When Retrieval Helps and Distracts: Evaluating Evidence-Generating LLMs for Biomedical Claim Verification
  3. 2026Linguistically-Aligned and Visually-Grounded Preference Optimization for Clinically-Augmented Medical Report Generation
  4. 2025VeriFact: Verifying Facts in LLM-Generated Clinical Text with Electronic Health Records
  5. 2026MedRAGChecker: Claim-Level Verification for Biomedical Retrieval-Augmented Generation
  6. 2026EHR-RobustGym: Benchmarking and Training Agents for Robust Clinical Reasoning
  7. 2026CliniCARE-Bench: Clinical Calibrated Audit of Medical Reasoning in EHR
  8. 2026Retrieval-Augmented LLMs for Evidence Localization in Clinical Trial Recruitment from Longitudinal EHR Narratives