DavidAU_L3.3_8B_OpusR_HUC
Babsie/DavidAU_L3.3_8B_OpusR_HUC
finetuneSAFETENSORS
Published metrics
- KL divergence
- 0.150
- Refusals
- 14/100
- Original refusals
- —
- Heretic version
- —
Published on the model card
Matched snippet
(Details: org censorship: 89/100 ; Heretic version: Refusals: 14/100, KL divergence: 0.1498 (less than 1 is excellent, 0=perfect) )
Provenance
- Canonical model
- allura-forge/Llama-3.3-8B-Instruct
- Declared parent(s)
- DavidAU/Llama3.3-8B-Instruct-Thinking-Heretic-Uncensored-Claude-4.5-Opus-High-Reasoning, allura-forge/Llama-3.3-8B-Instruct
- Classification
- derivative of a Heretic model without its own ablation evidence
- Created / updated
- 2026-03-05 / 2026-03-05
- Downloads
- 147 all time · 17 last 30 days
- License
- apache-2.0
Refusal rate 14%. Metrics are self-reported by the publisher; the directory does not re-run evaluations.