CMU-AIRe/TARS-7B
TEXT GENERATIONConcurrency Cost:1Model Size:7.6BQuant:FP8Ctx Length:32kPublished:Oct 24, 2025License:apache-2.0Architecture:Transformer0.0K Open Weights Cold

CMU-AIRe/TARS-7B is a 7.6 billion parameter open-source reasoning model developed by CMU-AIRe, based on Qwen2.5-7B-Instruct, with a 32K context length. It is specifically trained for safety using the TARS (Training Adaptive Reasoners for Safety) method, focusing on adaptive reasoning to achieve low refusal rates and safe behavior. This model is designed to facilitate research into reasoning models for LLM safety, particularly in handling harmful and harmless prompts.

Loading preview...