AI Paper Atlas

Research map / Vision-language and multimodal / Multimodal LLMs

OmniCoT: A Benchmark for Global and Multi-Step Panoramic Reasoning

Haocong He, Chenfei Liao, Zichen Wen, Zihao Dongfang, Xu Zheng, Bin Ren, Chang Su, Zixin Zhang, Harold Haodong Chen, Hongfei Zhang, Weijia Li, Kailun Yang and 4 more

ECCV 2026 · Vision-language and multimodal › Multimodal LLMs · approach: Spatial reasoning in VLMs · 0 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 SFT on OmniCoT-T to Qwen2.5-VL-7B (ablation)Qwen2.5-VL-7B-Instruct · OmniCoT-Boverall accuracy: 23.53 -> 49.31Improved
add GRPO stage after SFT (ablation)OmniCoT-R1 · OmniCoT-Boverall accuracy: 49.31 -> 59.54Improved
add GRPO after SFT (ablation)OmniCoT-R1 · OmniCoT-B Locateaccuracy: 39.44 -> 51.82Improved
add GRPO after SFT (ablation)OmniCoT-R1 · OmniCoT-BCoT F1: 38.07 -> 51.14Improved
use CoT instead of direct answer for open-source Qwen2.5-VL-72B (ablation)Qwen2.5-VL-72B-Instruct · OmniCoT-Boverall accuracy: 26.51 -> 35.26 (+8.75)Improved
use CoT instead of direct answer for LLaVA-v1.5-7B (ablation)LLaVA-v1.5-7B · OmniCoT-Boverall accuracy: 4.91 -> 13.77 (+8.86)Improved
use CoT instead of direct answer for LLaVA-v1.5-13B (ablation)LLaVA-v1.5-13B · OmniCoT-Boverall accuracy: 17.38 -> 15.03 (-2.35)Hurt
use CoT instead of direct answer for Gemini3-Flash (ablation)Gemini3-Flash · OmniCoT-Boverall accuracy: 39.97 -> 28.44 (-11.53)Hurt
use CoT instead of direct answer for GPT-5 (ablation)GPT-5 · OmniCoT-Boverall accuracy: 42.66 -> 39.72 (-2.94)Hurt
increase spatial anchors from 0 to 10 (ablation)InternVL3.5-14B · OmniCoT-Boverall accuracy: 28.35 -> 34.23Improved
increase spatial anchors from 0 to 10 (ablation)GPT-5 · OmniCoT-Boverall accuracy: 40.60 -> 44.03Improved
increase anchors from 0 to 3 (default) (ablation)Qwen2.5-VL-7B-Instruct · OmniCoT-Boverall accuracy: 25.79 -> 23.53Hurt

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