GitHub - THUDM/CaRR: This repository contains the code and data for the paper "Chaining the Evidence: Robust Reinforcement Learning for Deep Search Agents with Citation-Aware Rubric Rewards".

kept by eddie jan 14

Citation-aware rubric rewards (CaRR) and C-GRPO training improve deep search agents' reasoning quality and factual grounding over standard binary outcome rewards.

GitHub - Gen-Verse/OpenClaw-RL: OpenClaw-RL: Train any agent simply by talking
OpenClaw-RL is a fully asynchronous reinforcement learning framework that trains personalized AI agents from natural conversation feedback and supports real-world agentic RL across terminal, GUI, SWE, and tool-call settings.
also on Slime, #ai-agents, #open-source
AI infrastructure in the "Era of experience"
As open-source LLMs commoditize due to fierce Chinese competition, companies can build defensible moats through custom RL models trained on proprietary environments using LoRA and GRPO.
also on #reinforcement-learning, Qwen, #open-source
LLMs can invent their own compression - Rajan Agarwal
Co-training a summarizer and generator via RL lets an LLM learn its own text compression scheme, achieving ~9.5% context size with minimal prediction loss.
also on #reinforcement-learning, Qwen
AI's Plummeting Prices Are a Software Story, Not a Hardware One
The rapid decline in AI inference costs is driven by software innovations rather than hardware, enabling local open-weight models to compete with frontier APIs.
also on Qwen, #ai-agents, #open-source, #ai
2025: The year in LLMs
The 2025 LLM landscape was defined by reasoning models, coding agents, Chinese open weight models, prompt-driven image editing, CLI tools, and OpenAI losing its lead.
also on Qwen, #ai-agents, #open-source, #ai