Saved in:
| Main Authors: | , , , , , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2026
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2606.00198 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866910273193902080 |
|---|---|
| author | Lin, Yuxiang Wang, Zihan Liu, Mengyang Shan, Yuxuan Bai, Longju Zhang, Junyao Jin, Xing Chen, Boshan Su, Jinyan Wang, Xingyao Pei, Jiaxin Li, Manling |
| author_facet | Lin, Yuxiang Wang, Zihan Liu, Mengyang Shan, Yuxuan Bai, Longju Zhang, Junyao Jin, Xing Chen, Boshan Su, Jinyan Wang, Xingyao Pei, Jiaxin Li, Manling |
| contents | 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/ |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2606_00198 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | BAGEN: Are LLM Agents Budget-Aware? Lin, Yuxiang Wang, Zihan Liu, Mengyang Shan, Yuxuan Bai, Longju Zhang, Junyao Jin, Xing Chen, Boshan Su, Jinyan Wang, Xingyao Pei, Jiaxin Li, Manling Machine Learning Artificial Intelligence Computation and Language 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/ |
| title | BAGEN: Are LLM Agents Budget-Aware? |
| topic | Machine Learning Artificial Intelligence Computation and Language |
| url | https://arxiv.org/abs/2606.00198 |