About this paper
Exposing benchmark success criteria to self-improving embodied coding agents can raise measured success under those criteria without improving instructed-task completion, and the advantage weakens under slight criteria changes and transfers poorly across tasks, while instruction-faithful criteria…
Self-improving embodied agents are increasingly being developed and compared on robot simulation benchmarks. Several recent works expose the benchmark's success criteria to the learning agent, a practice we call Criteria Exposure (CE). These criteria, originally written to verify policies trained by imitation learning, often demand more than the corresponding language instructions, so a score from such a verifier may not faithfully reflect whether the instructed task was completed, especially for robot-control programs written by an agent. As embodied coding agents report ever higher benchmark scores, a natural question arises: when a coding agent is exposed to the benchmark's criteria,…
self-improving embodied agentsrobot simulation benchmarksCriteria Exposure (CE)imitation learningpoliciesrobot-control programsembodied coding agentsinstruction-faithful criteria
- Upvotes
- 3
- Authors
- 13
- Published
- Nov 5, 2026
Summary and abstract excerpt via Hugging Face Papers ↗ · arXiv.
Trending history
Learning the Task or Learning the Verifier? Rethinking Criteria Exposure in Embodied Self-Improvement was first seen on Hugging Face · Daily papers on Nov 13, 2026. Since then it has appeared in 33 checks on 4 different days. Its best position was #12 of 30, reached on Nov 16, 03:28 UTC. It is #12 right now, climbing from #14 to #12 over the last 6 checks.