Llama-3.1-SuperNova-8B-Lite_TIES_with_Base-Heretic
ChiKoi7/Llama-3.1-SuperNova-8B-Lite_TIES_with_Base-Heretic
hereticSAFETENSORS
Published metrics
- KL divergence
- 0.053
- Refusals
- 3/100
- Original refusals
- 97/100
- Heretic version
- v1.1.0
Published on the model card
Matched snippet
| Refusals | 3/100 | | KL divergence | 0.0530 |
Provenance
- Canonical model
- Joseph717171/Llama-3.1-SuperNova-8B-Lite_TIES_with_Base
- Declared parent(s)
- Joseph717171/Llama-3.1-SuperNova-8B-Lite_TIES_with_Base
- Classification
- card states the weights were abliterated with Heretic
- Created / updated
- 2025-12-22 / 2025-12-23
- Downloads
- 103 all time · 27 last 30 days
- License
- llama3.1
Published benchmarks
Copied from the model card. These never affect the ranking.
| Metric | This model | Original | Delta |
|---|---|---|---|
| IFEval | 80.96 | — | — |
| MATH | 15.56 | — | — |
| GPQA | 30.96 | — | — |
| MMLU-Pro | 38.8 | — | — |
Refusal rate 3%. Metrics are self-reported by the publisher; the directory does not re-run evaluations.