Harness Self-Evolution for Long-Horizon Terminal Agents
Long-horizon terminal agents may solve a task in one run yet fail across repeated runs. Stochastic errors can propagate through actions and context, creating a gap between capability and reliable execution. We present RiseHarness, a framework for iteratively improving agent harnesses with model weights fixed. It focuses on flipping tasks with both successful and failed runs under the same configuration, using successful runs as references for failure analysis. RiseHarness combines agent trajectories with internal harness traces from a runtime monitor to examine how model actions interact with command execution, feedback, context management, and completion decisions. A Harness Handbook lin...
| Authors | Ruhan Wang, Yucheng Shi, Zongxia Li et al. |
|---|---|
| Primary category | cs.AI |
| Published | 20 October 2026 |
| Last updated at source | 20 October 2026 |
Recent changes
- 21 Oct 2026 Description: Long-horizon terminal agents may solve a task in one run yet fail across repeated runs. Stochastic errors can propagate through actions and context, creating a gap between capability and reliable execution. We present RiseHarness, a framewo → Long-horizon terminal agents may solve a task in one run yet fail across repeated runs. Stochastic errors can propagate through actions and context, creating a gap between capability and reliable execution. We present RiseHarness, a framewo
- 21 Oct 2026 Topics added cs.AI, cs.CL removed hf-daily-papers, agent harnesses, runtime monitor, harness traces, behavioral checks, harness adaptation, execution evidence
Changes detected by Silytics between refreshes of the source.
Silytics shows public metadata from arXiv and links to the original. Information may be incomplete or out of date. Last refreshed 21 October 2026. Licenses and terms of the original apply to the work itself. Not financial, legal or professional advice.