MM-OptBench is a solver-grounded benchmark showing current multimodal LLMs reach at most 52% pass@1 on generating correct optimization models from text-plus-visual problem specifications.
hub Canonical reference
MM-Eureka: Exploring the Frontiers of Multimodal Reasoning with Rule-based Reinforcement Learning
Canonical reference. 83% of citing Pith papers cite this work as background.
50
Pith papers citing it
Background
83% of classified citations
abstract
DeepSeek R1, and o1 have demonstrated powerful reasoning capabilities in the text domain through stable large-scale reinforcement learning. To enable broader applications, some works have attempted to transfer these capabilities to multimodal reasoning. However, these efforts have been limited by the limited difficulty of selected tasks and relatively small training scales, making it challenging to demonstrate strong multimodal reasoning abilities. To address this gap, we introduce the MMK12 dataset and MM-EUREKA with 7B and 32B parameters. The former is a high-quality multimodal mathematics reasoning dataset featuring diverse knowledge domains with human-verified answers and solution processes. The latter is a multimodal model employing rule-based reinforcement learning on MMK12, utilizing online filtering and two-stage training strategy to enhance training stability. MM-EUREKA demonstrates remarkable performance gains in multimodal mathematical reasoning, outperforming previous powerful models like InternVL2.5-78B or InternVL2.5-38B-MPO. In particular, MM-EUREKA achieves competitive or superior performance compared to both open-source and closed-source models, and trails slightly behind o1 in multidisciplinary reasoning tasks. We open-source our complete pipeline to foster further research in this area. We release all our codes, models, data, etc. at https://github.com/ModalMinds/MM-EUREKA
hub tools
citation-role summary
background 9
baseline 1
dataset 1
method 1
citation-polarity summary
representative citing papers
S1-VL combines structured scientific reasoning with iterative image manipulation via code execution to reach state-of-the-art results on visual and scientific reasoning benchmarks.
Formalizes Reasoning Portability (RP) and proposes RDB-CL to modulate per-sample KL regularization in RLVR for MLLM continual learning, achieving +12.0% Last accuracy over vanilla RLVR baseline by preserving reusable reasoning on high-RP samples.
ATLAS uses a single functional token to unify agentic and latent visual reasoning without image generation or external execution.
CurveBench is a new benchmark for recovering rooted containment trees from images of nested Jordan curves, where the strongest model reaches only 19.1% accuracy on hard cases and fine-tuning lifts an open model to 33.3% on easy cases.
Audited olympiad corpus and Physics-R1 recipe improve 8B VLM by up to 18 points on held-out physics problems while exposing contamination in prior evals.
Omni-R1 unifies multimodal reasoning by generating intermediate images during the process in a SFT-plus-RL framework, with an Omni-R1-Zero variant that matches or exceeds it using only text data.
GPRO trains a meta-controller on 790k failure-labeled samples to dynamically select fast, perception, or reasoning paths in LVLMs, yielding higher accuracy and shorter responses than prior slow-thinking methods.
MGPO elicits grounding in LMMs via multi-turn RL with binary rewards, yielding 5.4% and 5.2% gains on MME-Realworld and V* Bench and surpassing GPT-4o on the latter after training on 21K samples.
GUI-R1 uses reinforcement fine-tuning with GRPO on a small curated dataset to create a generalist vision-language action model that outperforms prior GUI agent methods across mobile, desktop, and web benchmarks using only 0.02% of the data.
R1-VL uses StepGRPO with rule-based StepRAR and StepRVR rewards to let MLLMs learn step-by-step reasoning beyond imitation of positive paths.
RAPO uses an information-theoretic lower bound on visual gain to select high-entropy reflection anchors and optimizes a chain-masked KL surrogate, delivering gains over baselines on reasoning benchmarks across LVLM backbones.
SeePhys Pro benchmark reveals multimodal models degrade on physics reasoning as information transfers from text to images, with blind training improvements often stemming from textual cues rather than visual evidence.
CGC improves fine-grained multi-image understanding in MLLMs by constructing contrastive training instances from existing single-image annotations and adding a rule-based spatial reward, achieving SOTA on MIG-Bench and VLM2-Bench with transfer gains to other multimodal tasks.
DeepEyes uses reinforcement learning to teach vision-language models active perception and image-based thinking, yielding gains on perception, reasoning, grounding, and hallucination benchmarks.
ParaVT introduces the first multi-agent RL framework for parallel video tool calling in LMMs, using PARA-GRPO to resolve the Tool Prior Paradox and achieve +7.9% average improvement over Qwen3-VL baseline across six benchmarks.
DMPO approximates forward KL minimization in on-policy RL by aligning the policy to a group-level reward-proportional target distribution, yielding 9-12% relative gains over GRPO on NP-Bench and smaller gains on math reasoning.
VideoSeeker integrates agentic reasoning and visual prompts into LVLMs via automated data synthesis, cold-start supervision, and RL training, yielding +13.7% gains on instance-level video tasks over baselines including GPT-4o.
Data curation alone raises VLM accuracy by more than 11 points on average across many benchmarks while cutting required training compute by up to 87 times.
Attention dispersion during extended reasoning impairs MLLM perception on images, and a training-free VRGA framework mitigates it by selecting and reweighting visual attention heads using an entropy-focus criterion.
LongVT adds native video-cropping tool calling to LMMs for interleaved multimodal chain-of-tool-thought reasoning on long videos and releases VideoSIAH data for training and evaluation.
REVISOR adds multimodal visual-text reflection and a Dual Attribution Decoupled Reward to improve long-form video reasoning in MLLMs without extra supervised fine-tuning.
PAPO integrates perception-aware supervision via a KL-based loss into RLVR methods like GRPO, yielding 4.4-17.5% gains on multimodal benchmarks and 30.5% fewer perception errors, with larger gains on vision-heavy tasks.
v1 adds a point-and-copy mechanism for dynamic visual token referencing in multimodal reasoning, trained on a new 300K dataset with grounding annotations, and outperforms baselines on multimodal math tasks.