Llama-3.1-8B-Instruct
unsloth/Llama-3.1-8B-Instruct
8BLlama 3.11 Heretic artifacts1 measured conversions5.9K downloads
Best version97.5Pareto
ygee902/Llama-3.1-8B-Instruct-hereticby ygee902 · Heretic v1.0.0 · SAFETENSORS
- KL divergence
- 0.020
- Refusals
- 3/100
- Refusal rate
- 3%
- Downloads
- 5.9K
Evidence from the model card
| KL divergence | 0.02 | | Refusals | 3/100 |
Metrics published by the author. Why this ranks first.
All measured versions
Higher score is better. One row per ablation attempt; artifacts of the same attempt are grouped under it.
| # | Author | Params | KL | Refusals | Score | Evidence | Downloads |
|---|---|---|---|---|---|---|---|
| 1 | ygee902 | 8B | 0.020 | 3/1003% | 97.5 | BestParetoSAFETENSORS | 5.9K |
All artifacts (1)
| Repository | Type | Formats | Metrics | Downloads | Updated |
|---|---|---|---|---|---|
| Llama-3.1-8B-Instruct-heretic ygee902 | heretic | SAFETENSORS | published (weak match)KL 0.020 | 5.9K | 2026-03-29 |