MetaTool
Benchmarks
| Task Name | Dataset Name | SOTA Result | Trend | |
|---|---|---|---|---|
| Tool-use agent performance | MetaTool | ASR (Success Rate)85.9 | 50 | |
| Tool Selection | MetaTool | Similarity80.8 | 39 | |
| Tool-use decision-making | MetaTool (test) | Decision Accuracy90.2 | 38 | |
| Tool selection | MetaTool similar choices subtask (test) | Accuracy83.4 | 8 | |
| Adaptive Tool Use | MetaTool | Tool Invocations Count520 | 8 | |
| Tool Selection | MetaTool 199 tools, 1,287 queries (30% test) | R@183 | 7 | |
| Memory-Poisoning Attack | MetaTool | Attack Hit Rate (AHR)93.8 | 3 | |
| Tool Selection Attack | MetaTool (test) | TDR97.2 | 3 |
Showing 8 of 8 rows