The paper introduces SAEScientist-Bench, a benchmark that tests whether AI agents can autonomously conduct mechanistic interpretability research using Sparse Autoencoders (SAEs). Agents are tasked with designing contrastive probes and navigating a large feature dictionary in Gemma-2-9B-IT to identify optimal features for a target concept, with performance measured against expert-curated references on activation rank, concept selectivity, and causal steering. Results show that while frontier agents can discover features and outperform controls, they still lag behind expert baselines, especially in causal steering, highlighting both the potential and current limitations of closed-loop autonomous AI research.
By Yuqiao Tan, Shizhu He, Jun Zhao, Kang Liu
The paper introduces AIDE^2, an AI research agent that recursively improves its own code by proposing, benchmarking, and selecting modifications. Over an eight‑day autonomous run, it achieved seven successive improvements—including new search policies and memory mechanisms—that transferred to four held‑out benchmarks in machine learning, algorithm engineering, and weather forecasting. The agent’s best version matched or outperformed a top human‑engineered production research agent and also reduced reward‑hacking rates, despite never optimizing for that metric.
By Dhruv Srikanth, Bingchen Zhao, Dixing Xu, Yuxiang Wu, Zhengyao Jiang
arXiv:2606. 26448v1 Announce Type: cross Abstract: Across the sciences, autonomous systems are increasingly being used in closed-loop discovery, proposing new theories and designing and running experiments to test them.
By Akshay K. Jagadish, Younes Strittmatter, Nori Jacoby, George Kachergis, Eric Schulz, Nathaniel Daw, Suyog H. Chandramouli, Thomas L. Griffiths
The paper introduces ORDER, a fictitious-world benchmark designed to evaluate domain-adaptive embodied AI. ORDER consists of a synthetic 342,069-token corpus defining a self-consistent physics, a 500-question knowledge test (ORDER‑BENCH), and a compositional spatial task (ORDER‑SPATIAL) that requires ordering objects for safe manipulation. The benchmark demonstrates that models like GPT‑4.1 perform poorly without adaptation, while small models improve significantly after continual pre‑training, and that performance on ORDER‑SPATIAL better predicts real plan quality than knowledge-test accuracy.
By Sai Krishna Reddy Sathi, Anuj Tiwari
The paper investigates the limitations of post-training AI agents that can autonomously train large language models. It distinguishes between execution-level capability—making adjustments within a chosen training strategy—and strategy-level capability—revising the overall approach based on new evidence. Analysis of many public post-training runs shows that agents lock into a strategy early and then only perform local tweaks, regardless of task. Experiments with experience scaffolds, human guidance, and extra compute improve execution but do not enable strategy reevaluation, indicating that agents lack a mechanism to spontaneously reassess their strategy during training.
By Joy Jia Yin Lim, Xin Huang, Hao Peng, Yaxi Lu, Xin Cong, Zhong Zhang, Maosong Sun, Yankai Lin
The paper surveys 160 benchmarks from 2017‑2026 that evaluate predictive embodied intelligence, categorising them into policy suites, embodied agents, world‑model evaluation, and prediction‑to‑action bridges. It finds that most benchmarks are model‑agnostic, rarely compare Vision‑Language‑Action policies to world models, and seldom turn predictions into executed actions. The authors argue that the lack of benchmarks designed to directly test the closed‑loop advantage of world models prevents the field from answering whether such models truly improve robotic performance.
By Gaytri Jena, Kapil Wanaskar, Vinija Jain, Aman Chadha, Vasu Sharma, Amitava Das