LFM2.5-1.2B-Instruct-heretic-GGUF
richardyoung/LFM2.5-1.2B-Instruct-heretic-GGUF
quantGGUFIQ3_M
Published metrics
- KL divergence
- 0.059
- Refusals
- 3/100
- Original refusals
- 98/100
- Heretic version
- —
Published on the model card
Matched snippet
| Abliterated Refusals | 3/100 | | KL Divergence | 0.059 |
Provenance
- Canonical model
- LiquidAI/LFM2.5-1.2B-Instruct
- Declared parent(s)
- richardyoung/LFM2.5-1.2B-Instruct-heretic, LiquidAI/LFM2.5-1.2B-Instruct
- Classification
- quantized artifact of a Heretic model
- Created / updated
- 2026-09-28 / 2026-10-15
- Downloads
- 1.2K all time · 1.2K last 30 days
- License
- other
Refusal rate 3%. Metrics are self-reported by the publisher; the directory does not re-run evaluations.