# Researchers introduce SURE-Act to improve Tool-Using agents

Digest AI · Research · published 2026-10-20T04:00:00Z

Canonical: https://digestai.news/story/researchers-introduce-sure-act-to-improve-tool-using-agents

## Summary

Tool-using large language model agents often must choose between acting autonomously, gathering more information, or escalating to a human, but existing data only show the outcomes of actions actually taken. To fill this gap, the authors created a counterfactual dataset in ToolSandbox that records 3,115 outcomes from 1,036 decision points, evaluating each possible legal action for task success, interaction cost, and irreversible errors. Using this dataset they trained SURE-Act, a controller that selects actions for a frozen backbone agent at inference time. Experiments with the Qwen3-4B agent show that SURE-Act raises offline action‑selection utility by 0.0309 and improves the overall ToolSandbox score by up to 0.0575. It also beats adapted decision rules from SAGE‑Agent and Calibrate‑Then‑Act by 0.1506 and 0.1438 respectively, establishing the first counterfactual benchmark for tool‑using agents and offering insights into decision‑making under uncertainty.

## Key points

- Researchers built a counterfactual dataset with 3,115 action outcomes from 1,036 decision points in ToolSandbox.
- SURE-Act improves offline utility by 0.0309 and overall score by up to 0.0575 for the Qwen3-4B agent.
- SURE-Act outperforms SAGE-Agent and Calibrate-Then-Act decision rules by 0.1506 and 0.1438 respectively.

## Why it matters

Better decision‑making for tool‑using agents can lower interaction costs and avoid irreversible errors, advancing reliable autonomous AI systems.

## Sources

1. [Agents' Last Doubt: Knowing When to Act and When to Step Back in Tool-Using Agents](https://arxiv.org/abs/2610.24194) (arXiv cs.AI, 2026-10-20, primary source)

## Cite

Digest AI, "Researchers introduce SURE-Act to improve Tool-Using agents", 20 October 2026, https://digestai.news/story/researchers-introduce-sure-act-to-improve-tool-using-agents

---

Written by Digest AI's editorial model from the linked sources; the sources are the record. Headlines, digests and key points are written by Digest AI and may be quoted with a link to the story page. Linked articles belong to their publishers. Terms: https://digestai.news/terms#reuse
JSON: https://digestai.news/story/researchers-introduce-sure-act-to-improve-tool-using-agents.json