Front page

Earlier on the front page

Papers featured before, with the lines of inquiry they bear on. Newest first.

On the site·Featured Sep 27·Reinforcement Learning

Scaling Automatic Research Agents via World Models

Xiyuan Yang, Sheikh Sarwar, Jingru Cheng, Zhan Shi · 2026-08-12

Automating empirical research is a long-standing direction of AI. Recent automatic research (AutoResearch) agents bring this goal within reach, as modern LLMs show the capability to independently implement solutions and learn from the execution outcomes. Behind these gains, post-training (especially RL) plays a central role.

Research questions this paper bears on · from the site’s inquiring lines

arXiv · Excerpts and notes on the site

On the site·Featured Sep 27·Reasoning Architectures

BDH-CQ: In-Context Learning with Recurrent Latent Reasoning

Björn Engdahl, Adrian Kosowski, Jan Chorowski, Zuzanna Stamirowska · 2026-08-10

We introduce BDH-CQ, a reasoning model that combines in-context learning with recurrent latent reasoning. Inputs presented at inference time continuously update the model's recurrent memory; the model then solves a query through iterative computation in a high-dimensional latent space, without verbalizing its intermediate reasoning.

Research questions this paper bears on · from the site’s inquiring lines

arXiv · Excerpts and notes on the site