TLDRocket
Sign in

AI & Tools

2 summarised stories about AI & Tools, each linking back to the original source. Browse all topics →

Friday, 17 July 2026

The Tool Illusion: Rethinking Tool Use in Web Agents

arXiv cs.CL 6 hours ago

Researchers conducted a large-scale empirical study examining tool use in web agents, testing multiple tool sources, models, and evaluation benchmarks to clarify whether tools consistently improve performance. The study evaluated tools across diverse experimental settings and frameworks to establish comparable baselines that prior work had lacked. The findings provide clearer evidence about tool design principles and potential drawbacks, establishing a more reliable foundation for developing web agents that use tools.

Eta Given Delta: Defining LLM Tool Efficiency With Marginal Tool Utility

arXiv cs.CL 6 hours ago

Researchers introduced tool efficiency, a metric measuring the rate of useful tool calls in language model agent trajectories, alongside marginal tool utility to determine whether individual tools contribute value or can be removed. The paper uses LLM-as-a-Judge to evaluate each tool call and assess whether removing tools would maintain accuracy while improving efficiency. This metric enables direct measurement of tool efficiency in post hoc trajectory analysis, supporting future research on optimizing agent performance beyond accuracy.