Evidence-Grounded Reasoning

Follow this topic Ask about this topic

Evidence-grounded reasoning asks whether an AI conclusion follows from the documents, records, images, or tool outputs available to it, rather than merely sounding plausible. Common approaches retrieve and organize sources into evidence graphs or typed records, then verify claims, compare supporting and contradicting evidence, and calibrate answers or abstentions. Training increasingly rewards evidence use and citation faithfulness, while benchmarks test multi-hop integration, evidence removal or corruption, noisy contexts, and process quality—not just final-answer accuracy. A central challenge is ensuring evidence changes reasoning appropriately when sources are incomplete, conflicting, misleading, or costly to inspect, without causing excessive abstention.

Latest papers

91 papers13 papers in the last four weeks, down 19% on the four weeks before. 0.1% of all new papers.

Do Language Models Know What Compelling Scientific Evidence Looks Like?
Oct 29, 2026Conor Igoe, Gabriel Grand, Zane Koch +6Scientific Reasoning in Language Models Evidence-Grounded Reasoning
Edison Scientific Inc., San Francisco, CA, USA · MIT CSAIL, Cambridge, MA, USA
Identity or Evidence? Auditing Evidence Dependence in Agentic Life-Science Benchmarks
Oct 27, 2026Danai Brilli, Maria GeorgilaScientific Agent Benchmarks Evidence-Grounded Reasoning
Zo Labs
Trainable Boundary-Keeping Where No Formal Answer Exists: A Controlled Knock-In Study of Reward Training in Evidence-Sufficiency Judgment
Oct 26, 2026YaJie YinLLM Abstention Refusal Behavior in Language Models
Benefit-Targeted Routing and Bounded Revision for LLM-Based Evidence Screening in Manufacturing Supplier Discovery
Oct 19, 2026Yijiashun Qi, Hanzhe Guo, Yuxuan LiLLM Routing Evidence Selection
University of Michigan, Ann Arbor, MI 48109, USA · University of Pennsylvania, Philadelphia, PA 19104, USA
EGIP: An Evidence-Grounded Interaction Protocol for Clinical Agents
Oct 17, 2026Zecheng Ren, Ruichen Cong, Yi Min Chen +8Evidence-Grounded Reasoning Clinical Decision-Making
Waseda University, Tokorozawa, Japan · Shanghai Jiao Tong University, Shanghai, China
When Reasoning Traces Conflict with Evidence: How Answer Configuration Shapes Evidence Following
Oct 14, 2026Shaoang Li, Yanhang Shi, Jian LiReasoning Trace Analysis Evidence-Grounded QA
Stony Brook University
LabOPBench: A Multimodal Benchmark for Evidence-Grounded Laboratory Decision Making
Oct 14, 2026Jingyuan Li, Yujie Liu, Yijie Dai +18Evidence-Grounded Reasoning Scientific Reasoning in Language Models
Shanghai Artificial Intelligence Laboratory · Shanghai Jiao Tong University · Fudan University +6
JevGPT: Thinking Fast and Slow in Autonomous Research
Oct 13, 2026Yasheng Sun, Yanning Dai, Yimeng Chen +9Dual-Process Reasoning Autonomous Research Agents
King Abdullah University of Science and Technology (KAUST)
ACED: Reference-Anchored Aggregation of Natural-Language and Executable Evidence for Mathematical Reasoning
Oct 12, 2026Yiyi Lu, Jiaxiang Geng, Bing LuoLLM Mathematical Reasoning LLM Reasoning
1Duke Kunshan University · 2The University of Hong Kong
OWLET: On-Policy Weighted Learning via Evidence Transfer for Video Large Language Model Post-Training
Oct 9, 2026Qitan Lv, Ke Fan, Hong Wang +5Video-Language Models Large Vision-Language Models
University of Science and Technology of China · Shanghai AI Laboratory · Tsinghua University
From Evidence to Action: How Tool-Using Agents Fail
Oct 6, 2026Hongzhan Lin, Shidong Cao, Ziyang Luo +3Agent Failure Analysis Evidence-Grounded Reasoning
National University of Singapore · Hong Kong Baptist University · Amazon Web Services +1
Cite What You Explore: Budget-Aware LLM Reasoning over Medical KGs with Verifiable Evidence
Oct 6, 2026Chen Chen, Dongjie Wang, Mei Liu +1LLM Reasoning with Graphs Evidence-Grounded Reasoning
Electrical Engineering and Computer Science, University of Kansas, USA · Health Outcomes and Biomedical Informatics, University of Florida, USA
Proof-Grounded Patient-Specific Clinical Explanations from Knowledge-Graph Reasoning
Oct 5, 2026Surajit DasData Provenance Evidence-Grounded Reasoning
Calibrated Evidence Inference With a Small, Permissively Licensed Open Model
Oct 4, 2026Johann Emmanuel LiClinical Language Model Evaluation Evidence-Grounded Reasoning
Are We Measuring Scientific Intelligence? Rethinking the Evaluation of AI Scientists
Oct 4, 2026Kate Zhang, Yuante LiAI Agent Evaluation Scientific Agent Benchmarks
School of Computer Science Carnegie Mellon University Pittsburgh, PA 15213
GLoC-EHR: Evidence-Cited Clinical Reasoning over Global Context and Local EHR Events
Oct 1, 2026Chaiho Shin, Kwangsoo KimEvidence-Grounded Reasoning Clinical Outcome Prediction
Interdisciplinary Program of Medical Informatics, Seoul National University College of Medicine · Department of Transdisciplinary Medicine, Seoul National University Hospital · Center for Data Science, Healthcare AI Research Institute, Seoul National University Hospital +1
SEAR: Spoofing Evidence-Grounded Audio Reasoning Benchmark for Audio Language Models
Sep 30, 2026Rong Wan, Suliu Qin, Jiaxi Li +5Audio-Language Model Evaluation Audio Deepfake Detection
University of Surrey · Singapore University of Technology and Design · Guangxi University +1
EHR-RobustGym: Benchmarking and Training Agents for Robust Clinical Reasoning
Sep 30, 2026Yitong Qiao, Yancheng Jin, Lei Liu +4Evidence-Grounded Reasoning AI Agent Benchmarks
Zhejiang University · Ant Healthcare, Ant Group
PathAnchor: Path-Structured Evidence for Scientific Agents
Sep 30, 2026Qiuhui Chen, Jiafan Lu, Shuaimin Tang +5Evidence-Grounded Reasoning Scientific QA
School of Information Science and Engineering, East China University of Science and Technology · College of Intelligent Robotics and Advanced Manufacturing, Fudan University · Tencent
AffectReveal: Event-Grounded Emotion Recognition Beyond Visual Appearances
Sep 29, 2026Yihao Qian, Runhao Zeng, Sicheng Zhao +3Evidence-Grounded Reasoning Multimodal Emotion Recognition
South China University of Technology, Guangzhou, China · Shenzhen MSU-BIT University, Shenzhen, China · Tsinghua University, Beijing, China
Can LLMs Value the Right Evidence? Evidence-Value Misalignment in Dynamic Medical Diagnosis
Sep 28, 2026Kehua Feng, Yunsheng Lu, Yitong Qiao +5LLM Evaluation Evidence-Grounded Reasoning
Zhejiang University · Ant Healthcare, Ant Group
Evidence-Inference Reconstruction: When The Evidence Is Recalled But The Reasoning Goes Wrong
Sep 27, 2026Megan Diehl, Ser-Nam LimMulti-Hop QA Evidence-Grounded Reasoning
University of Central Florida
When Evidence Changes the Subject: Subject-Typed Claim Licensing for Learned Routing
Sep 27, 2026Jian Chen, Zixuan YuanEvidence-Grounded Reasoning Scientific Claim Verification
The Hong Kong University of Science and Technology (Guangzhou), China
Counterfactual Self-Evolving Agents for Evidence-Grounded Reasoning
Sep 26, 2026Xing Han, Yuxin Wang, Chen Chen +8Self-Evolving Agents Evidence-Grounded Reasoning
Johns Hopkins University · MIT · University of Texas at Austin +1
EviGraph: Proof-Carrying Selective Recommendation over Temporal Public-Service Knowledge Graphs
Sep 21, 2026Yixi Zhou, Sikun Wang, Lei Fan +1Recommender Systems Evidence-Grounded Reasoning
Hong Kong Baptist University · Tokyo University of Science · University of Illinois Urbana-Champaign +1
BioPhys-Bridge: A Benchmark for Interdisciplinary Scientific Reasoning in Physics-Grounded Biological Research
Sep 15, 2026Qingyang XuEvidence-Grounded Reasoning Scientific Reasoning in Language Models
Independent Researcher Shanghai, China
HypoKG: Evidence-Disciplined Biomedical Hypothesis Generation Beyond Endpoint Knowledge
Sep 10, 2026Dominic Okonkwo, Adetayo Okunoye, Ismailcem Budak ArpinarLLM Reasoning with Graphs Evidence-Grounded Reasoning
School of Computing University of Georgia
Sci-MMR: Benchmarking Multi-Step Evidence-Grounded Scientific Reasoning in Multimodal Agents
Sep 10, 2026Jiaqiang Li, Yajie Yang, Zhiheng Xi +15Evidence-Grounded Reasoning Scientific Reasoning
Fudan NLP Lab · Fudan NLP Group
Compute-Bounded Security Assurance - Coverage, Verification, and Response under Resource Constraints
Sep 7, 2026Jithin VG, Ditto PSAI Assurance Evidence-Grounded Reasoning
Bud Ecosystem
SCIRIGOR:Evaluating Open-Ended Scientific Analysis Beyond Final Scores
Sep 5, 2026Bowen Liu, Shuo Nie, Bodong Du +1Scientific Agent Benchmarks LLM Agent Evaluation
HKUST · Alibaba Group

Oct 29, 2026J/K move · Enter open · S save

Do Language Models Know What Compelling Scientific Evidence Looks Like?

Conor Igoe, Gabriel Grand, Zane Koch, Albert Bou+5

Edison Scientific Inc., San Francisco, CA, USA · MIT CSAIL, Cambridge, MA, USA

Scientific Reasoning in Language Models Evidence-Grounded Reasoning Confidence Estimation in Language Models Scientific Agent Benchmarks Belief Updating in LLMs Probability Elicitation from Language Models

Open paper AskPDF arXiv

Expert scientists excel at weighing empirical findings from multiple sources that vary in both statistical reliability and epistemic relevance. To what extent are language models, which are increasingly driving autonomous discovery systems, capable of such evidence deliberation? We introduce EDB-Bio, a scientific reasoning benchmark curated from real-world data in three biological domains: drug-perturbation transcriptomics, CRISPR screens, and human genome-wide association studies (GWAS). Each of EDB-Bio's 96k episodes presents a sequence of experimental results from published studies and asks the model to predict the probability that a target hypothesis is true. Out-of-the-box, open-weight and frontier models alike demonstrate systematic underconfidence and attain Brier scores comparable to a trivial base-rate forecaster. Post-training against strictly proper scoring rules significantly improves both calibration and discrimination. This calibration pays off downstream: in a sequential experiment selection task, our Ibis0 model is 2.5 times more efficient than the best frontier model. Throughout, quantifying evidence weights as Bayes factors reveals how models update their beliefs over time, whether they agree with expert biologists on relevance, and where they exhibit inconsistencies and biases. Our work extends calibration training and uncertainty analysis to a realistic scientific forecasting setting with applications to human cancer biology. We release all code, data, and trained models.