paper-with-me

Papers

BAGEN: Are LLM Agents Budget-Aware?

2026-05-29 · Yuxiang Lin, Zihan Wang, Mengyang Liu, Yuxuan Shan, Longju Bai, Junyao Zhang, Xing Jin, Boshan Chen, Jinyan Su, Xingyao Wang, Jiaxin Pei, Manling Li arxiv

While agents are increasingly spending more resources, today agent cost is mostly measured only after execution. A Budget-Aware Agent (BAGEN) should treat budget as an active control signal, rather than a passive cost metric. We first systematically define budget estimation as internal budgets (from agent computation) and external budgets (from agent actions). We then formalize budget-awareness as progressive interval estimation: at each step of a plan, an agent should predict an upper and lower bound on remaining budget, and alert when completion is unlikely. Scoring with a rollout-replay protocol, we find consistent failure patterns on four environments and five frontier agents: (1) strong agents do not necessarily have strong budget-awareness, with correlation r=0.35. (2) frontier models are consistently over-optimistic, continue spending on tasks that are unlikely to succeed, instead of alerting the user early. (3) budget-aware signal is actionable and trainable. Early stop saves 28-64% tokens on failed trajectories, and SFT+RL strengthens early stop and alert behavior. (4) precise interval calibration remains challenging, with interval coverage capping at 47% after SFT+RL. Project page: https://ragen-ai.github.io/bagen/

📄 PDF Abstract BibTeX arXiv:2606.00198

Code (0)

등록된 구현이 없습니다.

Similar Papers 제목 키워드 기반

FUSE : Failure-aware Usage of Subagent Evidence for MultiModal Search and Recommendation

2025-11-15 · Tushar Vatsa, Vibha Belavadi, Priya Shanmugasundaram, Suhas Suresha 외 arxiv

Multimodal creative assistants decompose user goals and route tasks to subagents for layout, styling, retrieval, and generation. Retrieval quality is pivotal, yet failures can arise at several stages: understanding user …

ClawArena-Team: Benchmarking Subagent Orchestration and Dynamic Workflows in Language-Model Agents

2026-06-30 · Kaiwen Xiong, Haonian Ji, Shi Qiu, Zeyu Zheng 외 arxiv

Production large language-model (LLM) agents are increasingly deployed not as lone problem-solvers but as managers: a main model creates specialized subagents, delegates work, and orchestrates their parallel, asynchronou…

Evaluating Long-Context Reasoning in LLM-Based WebAgents

2025-12-03 · Andy Chung, Yichi Zhang, Kaixiang Lin, Aditya Rawal 외 arxiv

As large language model (LLM)-based agents become increasingly integrated into daily digital interactions, their ability to reason across long interaction histories becomes crucial for providing personalized and contextu…

Harness-1: Reinforcement Learning for Search Agents with State-Externalizing Harnesses

2026-06-01 · Pengcheng Jiang, Zhiyi Shi, Kelly Hong, Xueqiang Xu 외 arxiv

Search agents are often trained as policies over growing transcripts: the model must decide how to search while also remembering what it has seen, which evidence is useful, which constraints remain open, and which claims…

Reinforcement Learning

AgentFactory: A Self-Evolving Framework Through Executable Subagent Accumulation and Reuse

2026-03-18 · Zhang Zhang, Shuqi Lu, Hongjin Qian, Di He 외 arxiv

Building LLM-based agents has become increasingly important. Recent works on LLM-based agent self-evolution primarily record successful experiences as textual prompts or reflections, which cannot reliably guarantee effic…