Expired50% confidenceFactTime unknown
The SciMDR-7B model achieved a score of 49.1 on the SciMDEval benchmark, an improvement of 29.3 points over the base model
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
7/2/2026
First Seen
Valid until: 9/30/2026(expired)
Sources
Related Claims
UnverifiedSAM3 scored 48.8 on the zero-shot segmentation test of the SA-V benchmark, surpassing the previous best score of 38.5, an improvement of more than 26%57% similarUnverified在覆盖九大维度共403道题的测试中，Muse-Glimmer-30B综合均分为76.04，比Corpus 3.6 27B的75.57高出0.47分57% similarUnverifiedHy3's score on the MRCR (Multi-turn Referential Coreference Resolution) benchmark jumped from 42.9% to 75.1%.56% similarUnverifiedGemma 4 31B achieved a composite capability score of 11,452, placing it at the forefront among models of the same tier.55% similarUnverifiedThe optimized Qwen 397B model introduced by Meta achieved a level close to Claude Ops 4.8 through a self-trace approach.55% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/57289API
curl https://kongchang.com/api/v1/knowledge/claims/57289MCP
get_claim(id=57289)