AI Paper Atlas

Research map / Benchmarks / VSR

VSR leaderboard

24 papers at ICML, NeurIPS, ICLR, CVPR, AAAI, ICCV and ECCV, and the journals TPAMI, IJCV and JMLR report results on VSR. Best accuracy: 84.3 (PixelReasoner-7B + Ours (visual entropy), CVPR 2026).

Numbers are as each paper reports them, read by AI from its results tables and not checked by hand. Papers can differ in setting (training data, model size, extra data); the topic column shows what each paper is about. Open the paper before you compare or quote a number.

VSR · accuracy

24 papers, each with its best reported number for its own method; higher is better.

#accuracy ↑MethodPaperTopic
184.3PixelReasoner-7B + Ours (visual entropy)
PixelReasoner-7B
Understanding and Mitigating Hallucinations in Multimodal Chain-of-Thought Models
CVPR 2026 · Table 2
Multimodal LLMs
283+ DiG
Qwen3-VL-8B-Thinking
DiG: Differential Grounding for Enhancing Fine-Grained Perception in Multimodal Large Language Models
CVPR 2026 · Table 1
Multimodal LLMs
381.6Attention Intervention
Qwen2-VL
Why Is Spatial Reasoning Hard for VLMs? An Attention Mechanism Perspective on Focus Areas
ICML 2025 · Appendix Table: Qwen2-VL spati
Multimodal LLMs
481.3POLIA
Qwen2.5-VL-7B
POLIA: Policy Optimization with Visual-Object-Level Intrinsic Advantage for Multimodal Reasoning
ICML 2026 · Table 1
Multimodal LLMs
581.14Qwen2-VL-7B + SFT + RoPE scaling
Qwen2-VL-7B
Reading Images Like Texts: Sequential Image Understanding in Vision-Language Models
ICLR 2026 · Table 2
Multimodal LLMs
678.21SpatialVTS (vision/text)
Qwen2-VL-7B
Enhancing Spatial Reasoning Through Visual and Textual Thinking
AAAI 2026 · Table 1
Multimodal LLMs
777OURO-7B
Qwen2-VL-7B-Instruct
OURO: A Self-Bootstrapped Framework for Enhancing Multimodal Scene Understanding
ICCV 2025 · Table 1
Multimodal LLMs
875.6VALOR
Qwen3-8B
No Labels, No Problem: Training Visual Reasoners with Multimodal Verifiers
ICLR 2026 · Table 1
Multimodal LLMs
972.9GRIT
Qwen2.5-VL-3B
GRIT: Teaching MLLMs to Think with Images
NeurIPS 2025 · Table 1
Multimodal LLMs
1072.3+ PGT
Llama-3-8B (instruction tuning, CLIP-ViT-L/14)
PGT: Procedurally Generated Tasks for improving visual grounding in MLLMs
ICML 2026 · Table 1
Multimodal LLMs
14 more papers on this leaderboard, the baselines reported on it and a CSV download are free with a scime account.
Sign in free to see all

Not affiliated with any conference, journal or the benchmark's authors. Corrections: contact@latenticlabs.com.