Unverified50% confidenceBenchmarkExact time
研究人员发现，仅将函数描述中的retrieve替换为fetch，在某些模型上会导致调用成功率下降约15%
1
Sources
50%
Confidence
Long-term
Relevance
7/8/2026
First Seen
Sources
Agent Skills规范：AI智能体技能标准化与可复用性实践
githubGitHub Trending7/3/2026
Related Claims
Unverified少样本方式在格式化输出、分类任务和规则遵循方面的幻觉率可降低40%-70%66% similarUnverified业界仿真精度评估通常要求吞吐量误差小于15%66% similarUnverifiedUnsloth can reduce memory usage by approximately 70% in some scenarios, and claims no loss of precision.66% similarUnverified在HotpotQA等多跳推理基准上，ReAct相比Act-only方案的任务完成率提升约10-15个百分点，同时幻觉率显著降低64% similarUnverifiedAnthropic's research shows that even mild alignment training can cause a 2–8 percentage point decline in model performance on certain reasoning-intensive tasks.63% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/257610API
curl https://kongchang.com/api/v1/knowledge/claims/257610MCP
get_claim(id=257610)