Lists (3)
Stars
A compact implementation of SGLang, designed to demystify the complexities of modern LLM serving systems.
Official PyTorch implementation for "ReFusion: A Diffusion Large Language Model with Parallel Autoregressive Decoding"
Agent S: an open agentic framework that uses computers like a human
Agent Memory Playground: AI Agent Memory Design & Optimization Techniques
ToolOrchestra is an end-to-end RL training framework for orchestrating tools and agentic workflows.
🔥[NeurIPS'25] DeepFund: Pilot for Your Next Fund Investment
Latent Collaboration in Multi-Agent Systems
Awesome LLM Self-Consistency: a curated list of Self-consistency in Large Language Models
This repository allows reproduction of Poetiq's record-breaking submission to the ARC-AGI-1 and ARC-AGI-2 benchmarks.
A collection of token reduction (token pruning, merging, clustering, etc.) techniques for ML/AI
Official repository for DR Tulu: Reinforcement Learning with Evolving Rubrics for Deep Research
Official implementation of paper "Think-at-Hard: Selective Latent Iterations to Improve Reasoning Language Models"
MiroMind Research Agent: Fully Open-Source Deep Research Agent with Reproducible State-of-the-Art Performance on FutureX, GAIA, HLE, BrowserComp and xBench.
Designing Multi-Agent Systems with Zero Supervision
LongProc: Benchmarking Long-Context Language Models on Long Procedural Generation
Tongyi Deep Research, the Leading Open-source Deep Research Agent
Official implementation of "Fast-dLLM: Training-free Acceleration of Diffusion LLM by Enabling KV Cache and Parallel Decoding"
dInfer: An Efficient Inference Framework for Diffusion Language Models
MCP Toolbox for Databases is an open source MCP server for databases.
[NeurIPS'25] dKV-Cache: The Cache for Diffusion Language Models
A simple, unified multimodal models training engine. Lean, flexible, and built for hacking at scale.
Official Repository for "Glyph: Scaling Context Windows via Visual-Text Compression"
Rigourous evaluation of LLM-synthesized code - NeurIPS 2023 & COLM 2024
MiniCPM4 & MiniCPM4.1: Ultra-Efficient LLMs on End Devices, achieving 3+ generation speedup on reasoning tasks
The official code implementation for "Cache-to-Cache: Direct Semantic Communication Between Large Language Models"