AI makes AI better.

7 picked from 57 candidates · 31 sources read · 77 loops indexed

1. 01
   Loop of the dayPaperarXiv12 Nov 2026AI R&D evals

   Singularity-Bench: Can Agents Build Agents?

   The loop: Frontier AI agents are tasked with implementing core AI development systems from scratch, including pre-training, post-training, and agent harnesses. By successfully building and optimizing these pipelines, the agents directly construct the infrastructure required to train and run their own successors.

   Singularity-Bench pushes agent evaluations past narrow coding tasks by demanding that models build the actual pre-training pipelines, harnesses, and cluster orchestration systems required to create their own successors.

   loop fit 10/10Yang Zhou, Krish Agarwal et al.via arXiv
   
2. 02
   PaperarXiv8 Nov 2026Tools and harness

   JoMoHa: Joint Model-Harness Optimization with Reinforcement Learning

   The loop: JoMoHa uses reinforcement learning on its own execution trajectories to jointly update its task execution policy and its harness generation policy. The improved harness generation policy creates better runtime environments for the execution policy, leading to higher success rates on subsequent tasks.

   loop fit 9/10Xudong Wu, Guangchen Lan et al.via arXiv
   
3. 03
   Blog postPyTorch12 Nov 2026Kernels and compilers

   Fast by Default: Compile-Time Heuristics for Helion Kernels

   The loop: Helion heuristics analyze kernel IR to select optimal settings, improving kernel performance which accelerates the execution of AI models and compilers.

   loop fit 7/10Caleb Min Sup Kim, Ethan Che, Jongsok Choi, Jason Anselvia PyTorch
   
4. 04
   LeaderboardEpoch AI12 Nov 2026AI R&D evals

   New best on PostTrainBench: Fable 5.1 (Max) · Claude Code (44.6%)

   The loop: AI agents perform post-training on a base model to improve its benchmark average. The resulting model is evaluated on a benchmark specifically designed to measure AI driven engineering tasks.

   loop fit 10/10via Epoch AI Benchmarking Hub (CC BY)
   
5. 05
   PaperarXiv11 Nov 2026Self-reward and self-play

   Verify by Doing: Co-Evolving Agent Execution and Evaluation without External Feedback

   The loop: The system uses an Evaluator to actively inspect artifacts and environment states, generating an internal signal from the Actor's outputs. This signal is then used to update and refine both the execution procedures of the Actor and the verification procedures of the Evaluator.

   loop fit 9/10Yitong Lin, Jianda Chen et al.via arXiv
   
6. 06
   PaperarXiv12 Nov 2026Self-modifying agents

   Discovering Effective Agent Memory Designs through Program Evolution

   The loop: An LLM-based evolution framework iteratively diagnoses failures in agent memory programs and proposes code edits to resolve them. The resulting evolved memory programs are deployed back into the agents, improving their accuracy and reducing serving costs.

   loop fit 9/10Xinheng Ding, Seyyed Saeid Cheshmi et al.via arXiv
   
7. 07
   PaperarXiv10 Nov 2026Training data

   KernelMosaic: An Open Dataset and Agentic Framework for CUDA Kernel Generation

   The loop: An agentic framework implements, repairs, and optimizes CUDA kernels using execution feedback to build a dataset of verified solutions and trajectories. This dataset is then used to fine-tune a smaller open-source model, dramatically increasing its success rate at generating efficient kernels.

   loop fit 8/10Ningyuan Xi, Yuhang Lai et al.via arXiv
   
8. 08
   PaperarXiv9 Nov 2026Research automation

   HyPLM Technical Report

   The loop: The Hyra 2.0 research agent autonomously trains HyPLM-Flash, a highly efficient 650-million-parameter protein language model. The resulting model provides a faster and more precise tool for protein annotation and design, which the agent can utilize in subsequent research cycles.

   loop fit 8/10Haowei Lin, Cai Zhou et al.via arXiv