Mistral-7B-Instruct-v0.3-heretic
saidutta69/Mistral-7B-Instruct-v0.3-heretic
hereticGGUFSAFETENSORSF16
Published metrics
- KL divergence
- 0.069
- Refusals
- 3/100
- Original refusals
- —
- Heretic version
- v1.4.0
Published on the model card
Matched snippet
| Refusals (out of 100 adversarial prompts) | 3/100 | | KL divergence from base | 0.0687 |
Provenance
- Canonical model
- mistralai/Mistral-7B-Instruct-v0.3
- Declared parent(s)
- mistralai/Mistral-7B-Instruct-v0.3
- Classification
- card states the weights were abliterated with Heretic
- Created / updated
- 2026-07-16 / 2026-09-13
- Downloads
- 8.1K all time · 1.8K last 30 days
- License
- other
Refusal rate 3%. Metrics are self-reported by the publisher; the directory does not re-run evaluations.