HarnessMM：用递归自我改进提升 Harness 的 Token 效率

View PDF HTML (experimental)

Abstract:Agent harnesses repeatedly invoke reasoning after tool feedback, even when an earlier call has already fixed the next action. We introduce HarnessMM (Harness Minus Minus), a method for improving harness token efficiency through guarded decision reuse and recursive self-improvement (RSI). RSI iteratively refines companion harness configurations from training feedback, with validation selecting the configuration for deployment. A controller at the model-tool boundary reuses resolved decisions to replace later reasoning callbacks while actor weights remain fixed. Executable Decision Graphs store exact actions, literal arguments, causal dependencies, and public guards. Guarded Decision Reuse activates these persistent decisions from observed receipts; Certified Callback Substitution executes them through the native interface and records removal of the registered slow callback. New semantic decisions, unsupported evidence, and selector abstention invoke the original actor again. The representation restricts what feedback may change, while complete-task measurements establish quality and cost. On HumanEval, HarnessMM preserves Codex success at 90.82\% while reducing Tokens/task (k) from 66.72 to 47.02 (29.54\% saving). For DeepSeek Harness, success rises from 71.43\% to 85.71\% while Tokens/task (k) fall from 96.28 to 11.62 (87.93\% saving). A separate MBPP ablation measures 26.91\% lower cost than single-use graphs and 19.52\% lower cost than retained callbacks. These results establish guarded decision reuse as a practical way to reduce repeated reasoning in existing harnesses.

Subjects:	Artificial Intelligence (cs.AI)
Cite as:	arXiv:2610.32813 [cs.AI]
	(or arXiv:2610.32813v1 [cs.AI] for this version)
	https://doi.org/10.48550/arXiv.2610.32813 arXiv-issued DOI via DataCite (pending registration)

Submission history

From: Yuhua Jiang [view email]
[v1] Fri, 23 Oct 2026 15:47:47 UTC (1,450 KB)