An agent harness is the code that organizes context, maintains state, and coordinates tool calls for a language model. We study how to improve the harness under a limited evaluation budget while keeping model weights fixed. Our method, MESH-Harness, organizes each harness into functional modules with explicit role-spec...
Zhi-Wei Shang, Yun-Chen Huo, Ming-Rong Gong et al.· 0 citations
The executable harness surrounding a GUI model determines how observations are assembled, actions are executed, and verification, recovery, and termination are controlled. Compared with harness optimization for non-GUI agents, automatically optimizing this harness poses three coupled challenges: reconciling model inten...
Ge-Yi Yang, Zi-Kun Qu, Xiang Li et al.· 0 citations
SERA (Scale-Equalized Rollout Allocation), which redistributes a fixed rollout budget to approximately equalize these finite-rollout scaling factors and introduces a multiplicity correction to remove the additional prompt weighting induced by heterogeneous rollout counts.
Zi-Hao Chen, Fan-Xiang Xiong, Hong-Ran Ren et al.· 0 citations
The proposed dynamic bipartite matching framework supports the advantages of decentralized matching with LLM-based behavioral modeling and online learning under incomplete information for economic simulation and computational social science research.
Wang-Xuan Fan, Xiao-Yu Nie, Zhou-Tian Shi et al.· 0 citations
This work proposes an offline-to-online combinatorial experts selection (CES) framework that leverages offline evaluation to warm-start model capability estimation and employs online learning to adapt to capability shifts and correct offline inaccuracies.
Jinkun Xu, Minghan Wang, Zhiyong Wang et al.· Proceedings of the 32nd ACM...· 0 citations
COBRA-Skills is introduced, an efficient framework that formulates skill optimization as budgeted sequential optimization over a dynamically evolving candidate space and remains robust to changes in the agent harness and performs effectively when the target model itself is used for skill generation and refinement.
Ping-Chen Lu, Xiang-Yi Wang, Xiang Li et al.· 0 citations
Sparse Probing and Outcome-calibrated Targets OPD is introduced, which addresses two coupled decisions, where to probe and what to distill, through an acquisition--exploration--exploitation procedure and demonstrates the effectiveness of SPOT in improving reasoning performance while balancing solution quality and cover...
Zi-Kun Qu, Min Zhang, Ming-Ze Kong et al.· 8 citations
The EXPonential-weight algorithm for prompt Optimization} (EXPO) is proposed to automatically optimize the task description and meta-instruction in the meta-prompt for LLM-based agents and is extended to additionally optimize the exemplars (i.e., history of interactions) in the meta-prompt to further enhance the perfor...
Ming-Ze Kong, Zhiyong Wang, Yao Shu et al.· arXiv.org· 7 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.
