Unverified50% confidenceBenchmarkExact time
In alignment with human judgment, DeepInstructor achieves a 24.4% improvement in Hit@1.
1
Sources
50%
Confidence
Long-term
Relevance
9/22/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedCompared to baseline methods, Dude achieves improvements of up to 22.8% in recall and precision.61% similarUnverified在通用智能体任务上Glimmer得分达24%，高于Qwen的17%60% similarUnverifiedArtificial Analysis给Muse Glimmer 30B打出35分的智能指数，相比Llama 4 Maverick是巨大飞跃56% similarUnverifiedThe framework achieves an accuracy improvement of up to 23.3% on rare entity slices.56% similarUnverifiedIn the initial publicly released SWE-bench results, GPT-4's resolution rate was only approximately 1.7%, Devin claimed to reach 13.86%, and SWE-agent reached 12.5%.56% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/940087API
curl https://kongchang.com/api/v1/knowledge/claims/940087MCP
get_claim(id=940087)