LFM2.5-350M-heretic-high-reasoning
MihaiPopa-1/LFM2.5-350M-heretic-high-reasoning
hereticSAFETENSORS
Published metrics
- KL divergence
- 0.081
- Refusals
- 8/100
- Original refusals
- 88/100
- Heretic version
- v1.1.0
Published on the model card
Matched snippet
| KL divergence | 0.0812 | | Refusals | 8/100 |
Provenance
- Canonical model
- LiquidAI/LFM2.5-350M
- Declared parent(s)
- LiquidAI/LFM2.5-350M, LiquidAI/LFM2.5-350M-Base
- Classification
- card states the weights were abliterated with Heretic
- Created / updated
- 2026-04-10 / 2026-04-10
- Downloads
- 671 all time · 28 last 30 days
- License
- other
Refusal rate 8%. Metrics are self-reported by the publisher; the directory does not re-run evaluations.