GLM-5.2 is live. #1 throughput on OpenRouter, pay-per-token on FriendliAI. Try it today ➜
mustard-curx
Available on FriendliAI
Run this model inference on single tenant GPU with unmatched speed and reliability at scale.
Model Details
Model Provider
mustard-curx
Model Tree
EleutherAI/gpt-j-6b
Input Modalities
Output Modalities
Supported Functionality
EleutherAI/gpt-j-6bResearch on vulnerability-oriented membership inference, machine unlearning, and secure code generation. This is a causal language model for code/text generation, not a security guarantee or a production security scanner.
The checkpoint is released for reproducibility. Evaluation metrics should be reported together with the exact benchmark version, selector construction, decoding settings, and Bandit/HumanEval environment used to produce them.
The model can still generate insecure, incorrect, or harmful code. Generated code must be reviewed and tested before use. Unlearning performance is specific to the supplied forget set and evaluation protocol.
This derivative checkpoint follows the license and use restrictions of its base model: apache-2.0.