Matched-budget fine-tunes on Banking77 and five more public workflows (prompt injection, SMS spam, emotion, request routing), plus two Aegis chat-moderation fine-tunes.
GitHubModeration & safety
Jev vs fine-tuned Laya: Banking77 goes from 51.3% to 79.4% with 1,001 examples
- Banking77 Laya
- 51.3% → 79.4%
- Banking77 Jev
- 80.0%