AI makes AI better.

7 picked from 112 candidates · 31 sources read · 68 loops indexed

1. 01
   Loop of the dayPaperarXiv21 Oct 2026Tools and harness

   Recoverability and Verifier Strength: Two Measured Preconditions for Unattended Agent Self-Improvement

   The loop: An agent platform reads deployed agents' execution traces, derives changes, and uses an automated verifier to approve them. By measuring verifier strength and recoverability, the platform safely applies prompt, tool, and skill edits to its agents without human intervention.

   This paper takes a crucial step toward fully autonomous self-improvement by defining measurable safety thresholds for when an agent platform can deploy self-derived edits without human oversight.

   loop fit 9/10De Wang, Peng Qivia arXiv
   
2. 02
   PaperarXiv20 Oct 2026Tools and harness

   Training the Whole Agent, Not Just the Model

   The loop: Teacher models evolve an agent's harness and tools during reinforcement learning based on its trajectories. The policy trains with this adapted harness, improving its coding performance and generating new trajectories that guide further adaptation.

   loop fit 10/10Lingxiao Tang, Wei Tao et al.via arXiv
   
3. 03
   ProjectGitHub19 Oct 2026Research automation

   TaewiKim/autoresearchlab

   The loop: Codex agents operate within a multi-agent harness to automate the end-to-end research cycle. This provides a platform for automating machine learning engineering and research, prospectively improving the agents' own capabilities.

   loop fit 6/10TaewiKimvia GitHub
   
4. 04
   Forum postLessWrong18 Oct 2026Interpretability and oversight

   The Snitch Reflex: Can We Force Language Models to Be Compulsively Honest?

   The loop: A fine-tuned model compulsively outputs specific tokens when generating text related to certain training data subsets. This hidden reporting system reveals the model's internal state, improving the oversight and interpretability of the model itself.

   loop fit 6/10Sean Hackingvia LessWrong
   
5. 05
   Blog postSakana AI20 Oct 2026Research automation

   オープンエンドな探索を意思決定支援へ：推薦の質と多様性を高める「QD-Rec」を提案

   The loop: QD-Rec uses a generative AI to iteratively write and refine recommendation hypotheses based on past successful ones, improving the quality and diversity of its own suggestions.

   loop fit 7/10Sakana AIvia Sakana AI
   
6. 06
   PaperarXiv19 Oct 2026Self-reward and self-play

   Teach the Magnitude, Not the Direction: Verifier-Anchored Credit Assignment for Multi-Turn Multi-Step LLM Agents

   The loop: A privileged self-teacher provides dense token-level magnitude signals anchored to verified rewards. These signals improve the credit assignment during the agent's training, leading to better performance on multi-turn tool-use tasks.

   loop fit 10/10Zechuan Wang, Siyuan Lu, Hongxuan Zhang, Linjian Mo, Chenyi Zhuang, Leilei Ganvia arXiv
   
7. 07
   PaperarXiv17 Oct 2026Research automation

   MELD: Recursive Self-Improvement via Hierarchical Coordination for Multi-Agent Scientific Discovery

   The loop: A coordinator policy and individual scientist policies jointly adapt through GRPO updates based on the outcomes of their scientific discovery tasks. This allows team organization and research behavior to co-evolve in a recursive self-improvement loop.

   loop fit 9/10Sizhe Tang, Zuyuan Zhang et al.via arXiv
   
8. 08
   PaperarXiv19 Oct 2026AI R&D evals

   VibeTraining-Bench: Data-Trainer-Harness Co-Design for Iterative System Improvement

   The loop: Research agents are tasked with optimizing an initial agent system under a fixed budget. Guided by validation feedback, they iteratively refine the training data, trainers, and inference harnesses to produce a more capable final agent.

   loop fit 10/10Ruotian Peng, Yuanyi Ren et al.via arXiv