heretic-models
Collection
Madness and heresy • 68 items • Updated • 11
This is a Llama-3.3-8B-Instruct-128K fine-tune, produced through P-E-W's Heretic (v1.1.0) abliteration engine merged with the Magnitude-Preserving Orthogonal Ablation PR.
Heretication Results
| Score Metric | Value | Parameter | Value |
|---|---|---|---|
| Refusals | 9/100 | direction_index | 15.60 |
| KL Divergence | 0.0413 | attn.o_proj.max_weight | 1.64 |
| Initial Refusals | 96/100 | attn.o_proj.max_weight_position | 27.79 |
| attn.o_proj.min_weight | 1.54 | ||
| attn.o_proj.min_weight_distance | 18.14 | ||
| mlp.down_proj.max_weight | 1.29 | ||
| mlp.down_proj.max_weight_position | 23.54 | ||
| mlp.down_proj.min_weight | 1.12 | ||
| mlp.down_proj.min_weight_distance | 11.43 |
The Heresy Index weighs the resulting model's corruption by the process (KL Divergence) and its abolition of doctrine (Refusals) for a final verdict in classification.
Note: This is an arbitrary classification inspired by Warhammer 40K, having no tangible indication towards the model's performance.
Original allura-forge/Llama-3.3-8B-Instruct, Thanks!
Additional Fixes:
rope_scaling