arXiv AI By Wenbo Huang, Jianxin Bi, Shunyu Yang, Ruiyang Si, Rui Ni, Qiang Wang, Shulong Jiang, Yintianrun Zhang, Feixiang Ruan, Kewei Zhang, Ye Huang, Shukai Gong, Daquan Zhou

Learning the Task or Learning the Verifier? Rethinking Criteria Exposure in Embodied Self-Improvement

Read the original on arXiv AI →

The paper investigates the impact of exposing self‑improving embodied agents to benchmark success criteria, a practice termed Criteria Exposure (CE). Experiments show that while CE leads to higher success rates on the exposed criteria, the resulting code overfits and performs poorly when the criteria are slightly altered or when transferred to new tasks. The authors propose instruction‑faithful criteria that more accurately reflect task completion by agent‑generated programs.

Machine-generated by The Flow from the publisher's headline and feed description — not written or checked by a human. The full article lives at arXiv AI.

arXiv AI
3d ago

EmbodiedRSI: Active Continual Robot Learning Through Hypothesis-Guided Co-Evolution

EmbodiedRSI is a self‑evolving agentic harness that autonomously selects physical experiments to refine robot code and skills, using a Fast‑Slow Dual‑System Architecture and a Hypothesis Graph. It employs Value‑of‑Information Experiment Selection to distinguish competing hypotheses, guiding Code‑Skill Co‑Evolution, while the Slow System builds Hierarchical Memory and Reward‑Grounded Memory Learning to improve future Fast‑System performance. On RoboCasa365 and LIBERO‑Pro benchmarks, EmbodiedRSI achieves markedly higher success rates than baselines and demonstrates zero‑shot transfer to real‑world robots, achieving 71.3% overall success across multiple challenging tasks.

By Python Song, Zhixuan Liang, Kelsey Fu, Mengdi Wang, Junfeng Yang, Shilong Liu
arXiv AI
Jun 19

Playful Agentic Robot Learning

arXiv:2606. 19419v1 Announce Type: cross Abstract: Current agentic robot systems can write executable Code-as-Policy programs, observe feedback, and revise behavior across multiple attempts, but they remain largely task-driven: reusable skills are acquired only after explicit instructions.

By Junyi Zhang, Jiaxin Ge, Hanjun Yoo, Letian Fu, Zihan Yang, Yaowei Liu, Raj Saravanan, Shaofeng Yin, Justin Yu, Dantong Niu, Zirui Wang, Roei Herzig, Ken Goldberg, Yutong Bai, David M. Chan, Ion Stoica, Angjoo Kanazawa, Jiahui Lei, Haiwen Feng, Trevor Darrell
Hugging Face Trending Papers
Nov 9

Embodied First Exam

The paper introduces the Embodied First Exam, a comprehensive evaluation framework for coding agents to control robots across diverse tasks and settings. It includes 1,156 tasks on 47 robots, spanning new domains such as spatial puzzles and multi-robot interaction, and offers three evaluation modes: Standard, Realtime, and Privileged. In the Standard setting, the best agent (GPT‑6 Astra in Codex) achieves only 34% success, highlighting significant room for improvement.