AI Paper Atlas

Research map / Vision-language and multimodal / Multimodal LLMs

Molecular Identifier Visual Prompting and Verifiable Reinforcement Learning for Chemical Reaction Diagram Parsing

Jiahe Song, Chuang Wang, Yinfan Wang, Hao Zheng, Bowen Jiang, Rui Nie, Xingjian Wei, Junyuan Gao, Yubin Wang, Bin Wang, Lijun Wu, Jiang Wu and 2 more

ECCV 2026 · Vision-language and multimodal › Multimodal LLMs · approach: Multimodal LLM benchmarks · 3 citations (OpenAlex)

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
use molecule identifiers (IdtVP) instead of overlaid indexed boxes (BIVP)Gemini 3 Pro zero-shot · RxnScribe-test Hybrid MatchF1: 45.27 -> 71.70Improved
use identifiers instead of raw-image BROSGemini 3 Pro zero-shot · RxnScribe-test Hybrid MatchF1: 3.90 -> 71.70Improved
IdtVP on small open models zero-shotQwen2.5-VL-7B · RxnCaption-15k-test Soft MatchF1: 19.16 (BIVP 15.6 vs IdtVP 19.2)No clear effect
add Re3-DAPO RL after SFTRxnID (IdtVP) · RxnCaption-15k-test Hybrid MatchF1: 61.2 -> 64.4Improved
add Re3-DAPO RL after SFTRxnCaption (BIVP) · RxnCaption-15k-test Hybrid MatchF1: 59.8 -> 64.1Improved
add Re3-DAPO RL after SFTRxnCaption (BIVP) · ScannedRxn Hybrid MatchF1: 51.0 -> 45.1Hurt
add Re3-DAPO RL after SFTRxnID (IdtVP) · ScannedRxn Hybrid MatchF1: 54.5 -> 56.3Improved
add Re3-DAPO RL after SFTRxnID (IdtVP) · RxnScribe-test Hybrid MatchF1: 74.6 -> 75.0No clear effect
use mixed Soft:Hybrid reward 1:1 rather than Soft-only (ablation)IdtVP+RL · RxnCaption-15k-test Hybrid MatchF1: 62.8 -> 64.4Improved
use mixed reward rather than Hybrid-only (ablation)IdtVP+RL · RxnCaption-15k-test Hybrid MatchF1: 64.4 vs 64.4No clear effect
use human ground-truth identifiers instead of Mid-Mapper detector (ablation)IdtVP+RL · RxnCaption-15k-test Soft MatchF1: 74.5 -> 73.6Hurt
IdtVP vs BIVP serialization noise after SFTSFT models · training-set image-level error rateerror rate %: 12.88 -> 4.66Improved

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