AI Paper Atlas

Research map / Large language models / AI and society

An Open-Ended Benchmark and Formal Framework for Adjuvant Research with MLLM

yi chen, Yu Zhang, Jian Xu, Hua Yue, Xinming Wang, Zequan Lyu, Xu-Yao Zhang, Wei Wei, Cheng-Lin Liu

ICLR 2026 · Large language models › AI and society · approach: Benchmarking and reproducibility practice

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
closed-source instead of open-source modelsMLLMs · Adjuvant open-ended Q&ALLM Score: 7.3 vs 6.4Improved
closed-source instead of open-source modelsMLLMs · Adjuvant open-ended Q&ASTS: 0.7263 vs 0.6922Improved
closed-source instead of open-source modelsMLLMs · Adjuvant open-ended Q&ABERTScore: 0.5656 vs 0.5504Improved
think models instead of inference modelsMLLMs · Adjuvant open-ended Q&ARS, IS, STSImproved
scale Qwen3 from 32B to 235BQwen3 · Adjuvant open-ended Q&ALLM Score: 7.6 vs 7.6No clear effect
fine-tune Qwen3 on general scientific data (ScienceOne)Qwen3 · Adjuvant open-ended Q&ALLM ScoreMixed
use DeepSeek-V3 instead of other models for hallucination rejectionDeepSeek-V3 · Adjuvant hallucination rejectionOverall HRR: LLM Score 7.8 vs HRR 2.10%Hurt
use domain-specific biomedical models instead of general MLLMsBioGPT-Large-PubMedQA · Adjuvant open-ended Q&ALLM Score: 1.9 vs 7.3Hurt

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy