Stars
🚀 Ultra Recipe for Training Long-Horizon Search Agents - matching frontier AI's search capability with a 20B model + stateful harness
feishu-cli 是一个功能完整的飞书开放平台命令行工具。它将飞书文档、知识库、电子表格、消息、日历、任务等操作封装为简洁的命令行接口，核心能力是 Markdown ↔ 飞书文档双向无损转换。
PDF Parser for AI-ready data. Automate PDF accessibility. Open-source.
Heuristic Learning Blog Post
The batteries-included agent harness.
A benchmark for evaluating AI agents on frontier ultra long-horizon auto research tasks.
A unified library for building, evaluating, and storing speculative decoding algorithms for LLM inference in vLLM
Claude Code is an agentic coding tool that lives in your terminal, understands your codebase, and helps you code faster by executing routine tasks, explaining complex code, and handling git workflo…
A fast and soft pattern search for trillion-scale corpora.
AgentIR is a retriever specialized for Deep Research agents.
A Heterogeneous Benchmark for Information Retrieval. Easy to use, evaluate your models across 15+ diverse IR datasets.
BrowseComp-Plus: A More Fair and Transparent Evaluation Benchmark of Deep-Research Agent (ACL 2026 Main)
AI agents running research on single-GPU nanochat training automatically
a recursive self-improving harness designed to help your agents (and future iterations of those agents) succeed on any task
PyTorch building blocks for the OLMo ecosystem
This API provides programmatic access to the AlphaGenome model developed by Google DeepMind.
Miles is an enterprise-facing reinforcement learning framework for LLM and VLM post-training, forked from and co-evolving with slime.
A Reproduction of GDM's Nested Learning Paper
😊 TPTT: Transforming Pretrained Transformers into Titans
Quick illustration of how one can easily read books together with LLMs. It's great and I highly recommend it.
[R]einforcement [L]earning from [M]odel-rewarded [T]hinking - code for the paper "Language Models That Think, Chat Better"
🌾 OAT: A research-friendly framework for LLM online alignment, including reinforcement learning, preference learning, etc.