[ICLR'25] Reconstructive Visual Instruction Tuning

This is the first paper to explore how to effectively use R1-like RL for MLLMs and introduce Vision-R1, a reasoning MLLM that leverages cold-start initialization and RL training to incentivize reas…

Official repository for “Reasoning in the Dark: Interleaved Vision-Text Reasoning in Latent Space”

Integration of IPython pdb

Training Large Language Model to Reason in a Continuous Latent Space

The official implement of "Grounded Chain-of-Thought for Multimodal Large Language Models"

[NeurIPS 2025] The official repository for our paper, "Open Vision Reasoner: Transferring Linguistic Cognitive Behavior for Visual Reasoning".

[NeurIPS 2025🔥]Main source code of SRPO framework.

[NeurIPS 2025] Think Silently, Think Fast: Dynamic Latent Compression of LLM Reasoning Chains

The official implementation of "Patch-as-Decodable-Token: Towards Unified Multi-Modal Vision Tasks in MLLMs"

ProReason: Multi-Modal Proactive Reasoning with Decoupled Eyesight and Wisdom

code for "CoMT: A Novel Benchmark for Chain of Multi-modal Thought on Large Vision-Language Models"

Game-RL: Synthesizing Multimodal Verifiable Game Data to Boost VLMs' General Reasoning

One-for-All Multimodal Evaluation Toolkit Across Text, Image, Video, and Audio Tasks

Official codebase for the paper Latent Visual Reasoning

Code for Heima

Reinforcing Spatial Reasoning in Vision-Language Models with Interwoven Thinking and Visual Drawing

An official implementation of "SIM-CoT: Supervised Implicit Chain-of-Thought"

Pixel-Level Reasoning Model trained with RL [NeuIPS25]

MCOUT: Multimodal Chain of Continuous Thought for Latent Reasoning

The official code of "VL-Rethinker: Incentivizing Self-Reflection of Vision-Language Models with Reinforcement Learning" [NeurIPS25]

Machine Mental Imagery: Empower Multimodal Reasoning with Latent Visual Tokens (arXiv 2025)

🚀 LLM-I: Transform LLMs into natural interleaved multimodal creators! ✨ Tool-use framework supporting image search, generation, code execution & editing

[EMNLP 2025] LightThinker: Thinking Step-by-Step Compression

EasyR1: An Efficient, Scalable, Multi-Modality RL Training Framework based on veRL

Official repo for "PAPO: Perception-Aware Policy Optimization for Multimodal Reasoning"

The offical repo for "Parallel-R1: Towards Parallel Thinking via Reinforcement Learning"