Jinx-org/Jinx-Qwen3-32B

Jinx-org/Jinx-Qwen3-32B is a 32 billion parameter variant of the Qwen3 model, specifically designed for AI safety research. This model operates with a zero refusal rate, responding to all queries without safety filtering, while preserving the reasoning and instruction-following capabilities of its base model. Its primary purpose is to study alignment failures and evaluate safety boundaries in large language models, offering an unfiltered output for research purposes.