ai4ai

A research framework for principled agent self-improvement under frozen evaluators and declared mutation boundaries, recording verifiable lineage to make it reproducible and auditable.

Updated weekly — until humanity achieves AI4AI. A survey-grounded catalog of 432 papers on long-horizon agents, benchmarks, harness design, and recursive self-improvement.

AI Systems as Scientific Instruments for Understanding the Mechanisms of Intelligence

The official repo for the paper "Making AI Scientists Auditable from Evidence to Claim"

[EMNLP26 Findings] Official repository for DataChef: Cooking Up Optimal Data Recipes for LLM Adaptation via Reinforcement Learning

[NeurIPS 2026 Meta-Agents] PostTrainBench⁰ — Can LLM agents automate LLM post-training without gradients?

Open-source recursive self-improvement system.

Curated, tagged, verified papers on recursive self-improvement (RSI), self-evolving AI/agents, and AI for AI research (AI4AI)