developer tools
Failure-Specific Feedback May Be Mostly Another Prompt
A controlled robot-code study separates retry gains from resampling, static guidance and feedback matched to the actual failure.
Summary
A controlled robot-code study separates retry gains from resampling, static guidance and feedback matched to the actual failure.
On the same 252 initial failures, Qwen2.5-Coder-7B recovers 85.3 percent within five template-guided retries, versus 68.3 percent for capped resampling. Yet feedback borrowed from other episodes and fixed task guidance produce similar point estimates, with no detected advantage from matching the coarse failure label. A rim-grasp task even worsens when the template prescribes an incompatible center grasp. The simulation study argues that self-repair evaluations need resampling, mismatched feedback and concise static guidance as baselines.
Why it matters
A controlled robot-code study separates retry gains from resampling, static guidance and feedback matched to the actual failure.
Limits and context
No additional limitation was separately recorded.
Key claims
A controlled robot-code study separates retry gains from resampling, static guidance and feedback matched to the actual failure.
Evidence: source-2026-11-05-015
Sources
- arXiv preprint 2611.06863, original v1arXiv · primary research
Corrections
No corrections have been recorded for this story.