Harness-Updating Capability

Entities in the same semantic neighborhood but without a typed relation to this one — candidates for new edges or unrecognized duplicates.

• Harness-updating capability is flat in base capability: models from different capability tiers produce harness updates that lead to surprisingly similar gainsclaim0.860
  First major claim of the paper, supported by narrow spread across evolvers and case study
• Harness-Updating Gain (Δupdate)method0.820
  Metric measuring harness-updating capability as the mean pairwise gain across an anchor agent set
• which models produce useful harness updates?question0.814
  First open question the paper sets out to answer through evolver-side analysis
• harness-updating is flat in base capability: models across capability tiers produce updates that yield similar gains, and even the Qwen3.5-9B evolver induces gains comparable to Claude Opus 4.6quote0.814
  Verbatim summary of first major finding from conclusion
• Harness-Benefit Capabilityconcept0.800
  The capability of a task-solving agent to benefit from updated harnesses during task solving
• which models actually benefit from updated harnesses?question0.775
  Second open question the paper sets out to answer through agent-side analysis
• Capability budget should be allocated to the task-solving agent rather than the evolver, since harness-updating varies by at most 3.1 pp across evolversclaim0.764
  Primary design recommendation derived from harness-updating flatness finding
• Harness Activation Failureconcept0.761
  A failure mode where weak-tier models fail to invoke relevant harness artifacts (e.g., skills) during task solving