AI Paper Atlas

Research map / Vision-language and multimodal / Multimodal LLMs

Constructing and Interpreting Digital Twin Representations for Visual Reasoning via Large Language Models and Reinforcement Learning

Yiqing Shen, Mathias Unberath

ECCV 2026 · Vision-language and multimodal › Multimodal LLMs · approach: Reinforcement learning for visual reasoning · 1 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
add format rewards (ablation)DT-R1 · JiTBench Level 2J: 0.412 -> 0.808Improved
add DAG validation reward (ablation)DT-R1 · JiTBench Level 2J: 0.687 -> 0.808Improved
add code execution reward (ablation)DT-R1 · JiTBench Level 2J: 0.724 -> 0.808Improved
add accuracy rewards (ablation)DT-R1 · JiTBench Level 2J: 0.523 -> 0.808Improved
add format rewards on top of task accuracy only (ablation)DT-R1 · JiTBench Level 2J: 0.298 -> 0.808Improved
add format rewards on training success rate (ablation)DT-R1 · JiTBench Level 2training success rate: 31.5% -> 94.2%Improved
add SAM2 specialist (ablation)DT-R1 · JiTBench Level 2J: 0.684 -> 0.808Improved
add OWLv2 specialist (ablation)DT-R1 · JiTBench Level 2J: 0.732 -> 0.808Improved
add DepthAnything specialist (ablation)DT-R1 · JiTBench Level 2J: 0.756 -> 0.808Improved
add Qwen2.5-VL specialist (ablation)DT-R1 · JiTBench Level 2J: 0.778 -> 0.808Improved
use Qwen2.5-7B backbone instead of DeepSeek-R1-Distill-Qwen-7B (ablation)DT-R1 · JiTBench Level 2J: 0.808 -> 0.798No clear effect
use PPO instead of GRPODT-R1 · JiTBench Level 2J: 0.808 -> 0.794No clear effect

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