1.5−3.0× lossless training or pre-training speedup. An off-the-shelf, easy-to-implement algorithm for the efficient training of foundation visual backbones.

[CVPR' 24] Toolkit for 360Loc: A Dataset and Benchmark for Omnidirectional Visual Localization with Cross-device Queries Resources

反蒸馏 Skill：清洗你被迫写的 Skill 文件，看起来完整，核心知识留给自己。Anti-distillation for employee Skills.

将冰冷的离别化为温暖的 Skill，欢迎加入数字生命1.0！Transforming cold farewells into warm skills? It's giving rebirth era. Welcome to Digital Life 1.0. 🫶

A clean, single-file PyTorch implementation of Attention Residuals (Kimi Team, MoonshotAI, 2026), integrated with Grouped Query Attention (GQA), SwiGLU feed-forward networks, and Rotary Position Em…

[ICCV 2025] Keypoint detection and description learned from image pairs only - no depth, no pose, no artificial augmentation required.

Implementation of RDD (CVPR 2025).

PyTorch implementation of MLP-Mixer: An all-MLP Architecture for Vision

[CVPR'26 ] Unblur-SLAM: Dense Neural SLAM for Blurry Inputs

🤗 The largest hub of ready-to-use datasets for AI models with fast, easy-to-use and efficient data manipulation tools

HGRN2: Gated Linear RNNs with State Expansion

[ICLR 2025 & COLM 2025] Official PyTorch implementation of the Forgetting Transformer and Adaptive Computation Pruning

empty repository for future papers

[NeurIPS 2024] Depth Anything V2. A More Capable Foundation Model for Monocular Depth Estimation

Official repository for the AAAI 2026 paper "VPHO: Joint Visual-Physical Cue Learning and Aggregation for Hand-Object Pose Estimation"

A High-Fidelity Simulation Platform for Aerial-Ground Collaborative Robotics Based on AirSim

A Quirky Assortment of CuTe Kernels

The AI Scientist-v2: Workshop-Level Automated Scientific Discovery via Agentic Tree Search

Claude Autoresearch Skill — Autonomous goal-directed iteration for Claude Code. Inspired by Karpathy's autoresearch. Modify → Verify → Keep/Discard → Repeat forever.

AI agents running research on single-GPU nanochat training automatically