GLM-5.3 is live. Run Z.ai's latest model on Friendli Model APIs. Try it today ➜
failspy
Available on FriendliAI
Run this model inference on single tenant GPU with unmatched speed and reliability at scale.
Model Details
Model Provider
failspy
Model Tree
Input Modalities
Output Modalities
Supported Functionality
Well, after my abliterated models, I figured I should cover all the possible ground of such work and introduce a model that acts like the polar opposite of them. This is the result of that, and I feel it lines it up in performance to a certain search engine's AI model series.
This is microsoft/Phi-3-mini-128k-instruct with orthogonalized bfloat16 safetensor weights, generated with a refined methodology based on that which was described in the preview paper/blog post: 'Refusal in LLMs is mediated by a single direction' which I encourage you to read to understand more.
This model has been orthogonalized to act more like certain rhymes-with-Shmemini models.