Unverified50% confidenceBenchmarkExact time
After fine-tuning, Laya's accuracy on the MASSIVE dataset can be improved from 0.45 to 0.85, and on another task from 0.362 to 0.766.
1
Sources
50%
Confidence
Long-term
Relevance
9/25/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedVERGE improved precision from 0.764 to 0.849 and Matthews Correlation Coefficient from 0.681 to 0.73070% similarUnverifiedIn tests on 2,456 held-out prompts, increasing the sampling steps from 20 to 50 improved FID only from 27.0 to 26.6 (a gain of 0.4), while removing the shift caused FID to rise from 27.0 to 27.3 and FD-DINOv2 to rise from 218 to 228.65% similarUnverifiedThe composite score of reports selected by the full composite objective rose to 0.4122.65% similarUnverifiedExperiments found that the task_score under the shaped reward configuration was only 0.125, compared to 0.417 for the naive reward configuration, with an effect size of d=4.47.64% similarUnverifiedLaya achieves an accuracy of 0.892 on Yes/No binary true-or-false judgment tasks.63% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/950935API
curl https://kongchang.com/api/v1/knowledge/claims/950935MCP
get_claim(id=950935)