Unverified50% confidenceBenchmarkExact time
The researchers validated CITA on three tool-use benchmarks in combination with multiple different backbone LLMs, and the results show consistent improvements on both Tool F1 and task success rate metrics.
1
Sources
50%
Confidence
Long-term
Relevance
10/5/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedTool F1 is a tool-call quality metric that combines precision and recall, and is more fine-grained than task success rate alone.74% similarUnverifiedTesting tool-calling accuracy requires attention to three levels: tool selection accuracy, parameter extraction accuracy, and tool chain orchestration rationality, of which parameter extraction is often the most error-prone.73% similarUnverifiedThe research team validated the effectiveness of S3KG on nine benchmarks.72% similarUnverified标准基准测试衡量的是单次工具调用的格式正确性，而真实的生产级Agent失败往往发生在多步序列之中72% similarUnverifiedExperiments show that the gated objective achieved the highest average accuracy on both datasets, verifying that objective reliability and evaluation priority can be optimized independently71% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/975634API
curl https://kongchang.com/api/v1/knowledge/claims/975634MCP
get_claim(id=975634)