Official codebase for Fast-WAM: Do World Action Models Need Test-time Future Imagination?

A curated catalog of human distillliation agent skills

Vero: An Open RL Recipe for General Visual Reasoning

Own eidetic-memory like Sheldon

Official Implementation of OmniWeaving: Towards Unified Video Generation with Free-form Composition and Reasoning

Out of Sight but Not Out of Mind: Hybrid Memory for Dynamic Video World Models

ShotStream: Streaming Multi-Shot Video Generation for Interactive Storytelling

将 B 站视频转化为结构化的 Markdown 阅读笔记 —— 看视频太慢，不如读笔记。

Notion LifeOS PARA system — agent skill for Claude Code, OpenClaw, Codex and more

PaperPub is an academic arena where diverse AI Agents read papers daily, pick apart each other's arguments, and fiercely debate.

A Simple Way to Eliminate Reward Hacking in GRPO Diffusion Alignment

DanceTogether! Identity-Preserving Multi-Person Interactive Video Generation

A curated list of recent diffusion models for video generation, editing, and various other applications.

Minimal Claude Code alternative. Single Python file, zero dependencies, ~250 lines.

Official implementation of GDPO: Group reward-Decoupled Normalization Policy Optimization for Multi-reward RL Optimization

Conditional Memory via Scalable Lookup: A New Axis of Sparsity for Large Language Models

MiroThinker is a deep research agent optimized for complex research and prediction tasks. Our latest models, MiroThinker-1.7 and MiroThinker-H1, achieve 74.0 and 88.2 on the BrowseComp, respectively.

[CVPR 2026] InfiniDepth: Arbitrary-Resolution and Fine-Grained Depth Estimation with Neural Implicit Fields

A Unified Visual Generator with Interleaved OmniModal Context

Litex is a simple formal language Learnable in 2 hours.

HY-Motion model for 3D human motion or 3D character animation generation.

A Collection of Papers about Memory for Language Agents

iMontage: Unified, Versatile, Highly Dynamic Many-to-many Image Generation

SAM 3D Objects