Stars
[NeurIPS 2025] Implementation for the paper "The Surprising Effectiveness of Negative Reinforcement in LLM Reasoning"
Jacobi Forcing: Fast and Accurate Diffusion-style Decoding
The paper list of "Memory in the Age of AI Agents: A Survey"
[EMNLP'23, ACL'24] To speed up LLMs' inference and enhance LLM's perceive of key information, compress the prompt and KV-Cache, which achieves up to 20x compression with minimal performance loss.
slime is an LLM post-training framework for RL Scaling.
EasyR1: An Efficient, Scalable, Multi-Modality RL Training Framework based on veRL
LLaDA2.0 is the diffusion language model series developed by InclusionAI team, Ant Group.
A Survey of Reinforcement Learning for Large Reasoning Models
GoatWu / Self-Forcing-Plus
Forked from guandeh17/Self-ForcingUnofficial extension implementation of Self-Forcing to support I2V && 14B training.
Accelerating Large-Scale Reasoning Model Inference with Sparse Self-Speculative Decoding
Official Repository for "Glyph: Scaling Context Windows via Visual-Text Compression"
Latent Collaboration in Multi-Agent Systems
[Arxiv 2025] In-Video Instructions: Visual Signals as Generative Control
Kandinsky 5.0: A family of diffusion models for Video & Image generation
Video-as-Answer: Predict and Generate Next Video Event with Joint-GRPO
[ASPLOS'26] Taming the Long-Tail: Efficient Reasoning RL Training with Adaptive Drafter
Official code implementation of Context Cascade Compression: Exploring the Upper Limits of Text Compression
Official Implementation of "MMaDA-Parallel: Multimodal Large Diffusion Language Models for Thinking-Aware Editing and Generation"
[NeurIPS 2025 Oral]Infinity⭐️: Unified Spacetime AutoRegressive Modeling for Visual Generation
dInfer: An Efficient Inference Framework for Diffusion Language Models
Tiny Model, Big Logic: Diversity-Driven Optimization Elicits Large-Model Reasoning Ability in VibeThinker-1.5B
The code for NeurIPS 2025 paper "A-MEM: Agentic Memory for LLM Agents"
StreamingVLM: Real-Time Understanding for Infinite Video Streams