[ACL 2026 Best Paper Candidate (SAC Highlight / Oral)] Official implementation of "Double: Breaking the Acceleration Limit via Double Retrieval Speculative Parallelism".

A collection of useful Github repositories. Github项目精选。

AcadHomepage: A Modern and Responsive Academic Personal Homepage

Official Repository of Orchestra-o1: Omnimodal Agent Orchestration

将博导十年科研经验炼化为可直接调用的 AI 技能。从 Idea 构思到论文投稿，你的 AI 科研副导师。

🎵 将音乐搬进AI终端 · 打造Coding专属智能DJ · 听歌新范式 · 交互式音乐 ｜A retro pixel DJ for Claude Code / Codex CLI / Terminal with Apple Music / QQ Music / Local Music — plays music, shows synced lyrics, and 😱 panics when…

This is the official repo for the paper "General365: Benchmarking General Reasoning in LLMs under High Difficulty and Diversity".

This is the official repo for the paper "AMO-Bench: Large Language Models Still Struggle in High School Math Competitions".

Official code for "Self-Distilled Agentic Reinforcement Learning"

Official code for "KnowU-Bench: Towards Interactive, Proactive, and Personalized Mobile Agent Evaluation"

Official code for "SKILL0: In-Context Agentic Reinforcement Learning for Skill Internalization"

主要记录大语言大模型（LLMs） 算法（应用）工程师相关的知识及面试题

Extremely Long-Horizon Agentic Tasks Requiring Active Acting and Inductive Reasoning

ICLR2026 SAFER: Risk-Constrained Sample-then-Filter in Large Language Models

SafeGround: Know When to Trust GUI Grounding Models via Uncertainty Calibration

😼 优雅地使用基于 clash/mihomo 的代理环境

Draft-Target Disaggregation LLM Serving System via Parallel Speculative Decoding.

Official Repo for Open-Reasoner-Zero

Sky-T1: Train your own O1 preview model within $450

Minimal reproduction of DeepSeek R1-Zero

Official code for the paper, "Stop Summation: Min-Form Credit Assignment Is All Process Reward Model Needs for Reasoning"

s1: Simple test-time scaling

🔍 Search-o1: Agentic Search-Enhanced Large Reasoning Models [EMNLP 2025]

[NeurIPS 2024] The official implementation of paper: Chain of Preference Optimization: Improving Chain-of-Thought Reasoning in LLMs.

[NIPS'25 Spotlight] Mulberry, an o1-like Reasoning and Reflection MLLM Implemented via Collective MCTS

[CVPR 2024 Oral] InternVL Family: A Pioneering Open-Source Alternative to GPT-4o. 接近GPT-4o表现的开源多模态对话模型

[NeurIPS'23 Oral] Visual Instruction Tuning (LLaVA) built towards GPT-4V level capabilities and beyond.