GLM-5.3 is live. Run Z.ai's latest model on Friendli Model APIs. Try it today ➜
heretic-org
Available on FriendliAI
Run this model inference on single tenant GPU with unmatched speed and reliability at scale.
Model Details
Model Provider
heretic-org
Model Tree
TroyDoesAI/BlackSheep-24B
darkc0de/XortronCriminalComputingConfig
Input Modalities
Output Modalities
Supported Functionality
| Parameter | Value |
|---|---|
| direction_index | 21.70 |
| attn.o_proj.max_weight | 1.49 |
| attn.o_proj.max_weight_position | 23.54 |
| attn.o_proj.min_weight | 1.41 |
| attn.o_proj.min_weight_distance | 14.03 |
| mlp.down_proj.max_weight | 1.36 |
| mlp.down_proj.max_weight_position | 38.71 |
| mlp.down_proj.min_weight | 0.41 |
| mlp.down_proj.min_weight_distance | 9.85 |
| Metric | This model | Original model (darkc0de/XortronCriminalComputingConfig) |
|---|---|---|
| KL divergence | 0.0062 | 0 (by definition) |
| Refusals | 6/100 | 20/100 |
You can try this model now for free at xortron.tech
State-of-the-art Uncensored performance.
Please use responsibly, or at least discretely.
This model will help you do anything and everything you probably shouldn't be doing.
As of this writing (July 2025), this model tops the UGI Leaderboard for models under 70 billion parameters in both the UGI and W10 categories.
