Beta
Explore
Marketplace
Neural Labs
Chat
Wallet
Docs
Qwen3.5-35B-A3B-heretic-v3 – AI Model by meangrinch | AlphaNeural AI
You can deploy this model and start earning money today!
meangrinch
/
Qwen3.5-35B-A3B-heretic-v3
like
0
safetensors
qwen3_5_moe
heretic
abliteration
uncensored
en
unsloth/Qwen3.5-35B-A3B
finetune
apache-2.0
us
Views
No views yet
Model card
Files and Versions
Community
API
Deploy
Qwen3.5-35B-A3B-heretic-v3
Abliterated (uncensored) weights generated with an unreleased version of
Heretic
using the experimental Arbitrary-Rank Ablation (ARA) method.
See
https://github.com/p-e-w/heretic/pull/211
for details about ARA.
Abliteration parameters
Parameter
Value
start_layer_index
8
end_layer_index
29
preserve_good_behavior_weight
0.4053
steer_bad_behavior_weight
0.0021
overcorrect_relative_weight
1.0113
neighbor_count
15
Abliteration details
Metric
Value
Refusals
0/100
KL divergence
0.2623
A lower refusal count means the model is more willing to engage with restricted prompts.
A lower KL divergence means the abliterated weights deviate less from the original model's distribution (i.e. less capability degradation).
Source
Base model:
unsloth/Qwen3.5-35B-A3B
Related
GGUF quantized version:
Qwen3.5-35B-A3B-heretic-v3-GGUF
