analiza aiqbench
This research introduces ARC, a hierarchical policy that dynamically optimizes LLM agent configurations by treating them as semi-Markov decision processes rather than relying on static templates. By adapting workflows and resource budgets to specific query difficulties, the method significantly boosts reasoning accuracy and tool-use success rates while reducing wasted compute. The study validates that learning per-query configurations is a robust alternative to "one size fits all" designs for complex agentic systems.
Learning to Configure Agentic AI Systems
arXiv:2602.11574v4 Announce Type: replace Abstract: Configuring LLM-based agent systems involves choosing workflows, tools, token budgets, and prompts from a large combinatorial design space, and is typically handled today by fixed templates or hand-tuned heuristics that apply the same configuration regardless of query difficulty, leading to brittle behavior and wasted compute. To address this, we formulate agent configuration as a semi-Markov decision process (SMDP) where each configuration acts as a temporally extended option that determines how an agent system processes a query, and introduce introduce ARC (Agentic Resource & Configuration learner), a lightweight hierarchical policy that dynamically selects query-specific agent configurations. Across reasoning, tool-use, and agentic benchmarks, ARC consistently improves over budget-matched tool-augmented LLMs, increasing average reasoning accuracy by 31.3%, tool-use accuracy by 13.95%, and doubling {\tau}-Bench (Airline) Pass^1 success from 9.0% to 18.0%. These results demonstrate that learning per-query agent configurations is a powerful alternative to "one size fits all" designs.
Abstract:Configuring LLM-based agent systems involves choosing workflows, tools, token budgets, and prompts from a large combinatorial design space, and is typically handled today by fixed templates or hand-tuned heuristics that apply the same configuration regardless of query difficulty, leading to brittle behavior and wasted compute. To address this, we formulate agent configuration as a semi-Markov decision process (SMDP) where each configuration acts as a temporally extended option that determines how an agent system processes a query, and introduce introduce ARC (Agentic Resource & Configuration learner), a lightweight hierarchical policy that dynamically selects query-specific agent configurations. Across reasoning, tool-use, and agentic benchmarks, ARC consistently improves over budget-matched tool-augmented LLMs, increasing average reasoning accuracy by 31.3%, tool-use accuracy by 13.95%, and doubling {\tau}-Bench (Airline) Pass^1 success from 9.0% to 18.0%. These results demonstrate that learning per-query agent configurations is a powerful alternative to "one size fits all" designs.
Artificial Intelligence (cs.AI)
arXiv:2602.11574 [cs.AI]
(or arXiv:2602.11574v4 [cs.AI] for this version)
https://doi.org/10.48550/arXiv.2602.11574 Focus to learn more arXiv-issued DOI via DataCite
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Text integral sub licența deschisă a publicației · 238 cuvinte
