Unverified50% confidenceBenchmarkExact time
LinguaArray increased the number of solved instances by 105% to 397% relative to the corresponding backend solvers.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
10/30/2026
First Seen
Valid until: 1/28/2027
Sources
Related Entities
Related Claims
UnverifiedLinguaArray solved at most 69.1% of the instances.64% similarUnverifiedProviding specific input and expected output examples in the prompt can improve docstring generation quality from approximately 30% usable to approximately 85% usable.60% similarUnverifiedReAct 相比单纯的 Chain-of-Thought 提示，在多跳推理和信息检索任务上的成功率提升超过 30%58% similarUnverifiedUnder the best prompt, SemantEq-ProofAudit raises the recall on 407 equivalent statement pairs from 65.4% to 81.1%.57% similarUnverifiedAppending 1-5 real invocation examples to tool descriptions can raise tool-calling accuracy on certain benchmarks from about 72% to 90%.57% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1114067API
curl https://kongchang.com/api/v1/knowledge/claims/1114067MCP
get_claim(id=1114067)