Unverified50% confidenceBenchmarkExact time
The unadapted original backbone model scores 26.50 ANLS on SquintBench, the best general model (original view) scores 24.44, and the best off-the-shelf model under the downsampled view scores 68.60.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
10/12/2026
First Seen
Valid until: 1/10/2027
Sources
Related Entities
Related Claims
UnverifiedOn the Artificial Analysis Intelligence Index, the top six positions are entirely occupied by Anthropic's models, with Claude Opus 5.5 at the top, and Mistral Large 4 Preview ranked 25th out of 25 models.56% similarUnverifiedA model with approximately 27B parameters achieved a score of 52 running on a laptop, demonstrating that high-quality models are penetrating edge and lightweight devices.56% similarUnverifiedThe fine-tuned 406M model achieved 36.33 ROUGE-1 on the test set, while the authors' own 1.2B system achieved 35.41 ROUGE-1, and the two cannot be statistically distinguished.56% similarUnverifiedThe open-sourced model has a total of 305B parameters and is distributed in the form of 48 weight shards.54% similarUnverified在SWE-bench Pro基准上，千问3.8 27B模型的分数超过了Claude Opus54% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1001172API
curl https://kongchang.com/api/v1/knowledge/claims/1001172MCP
get_claim(id=1001172)