ccharnkij/Llama-3.1-8B-Instruct-Abliterated
TEXT GENERATIONConcurrency Cost:1Model Size:8BQuant:FP8Ctx Length:32kPublished:Nov 30, 2025License:llama3.1Architecture:Transformer Cold

ccharnkij/Llama-3.1-8B-Instruct-Abliterated is an 8 billion parameter instruction-tuned causal language model based on Meta's Llama 3.1 architecture, with a 32768 token context length. This model has undergone "abliteration" to remove refusal mechanisms, allowing it to comply with prompts that the original Llama 3.1-8B-Instruct would typically refuse. It is specifically designed for exploring LLM internals and mechanistic interpretability, demonstrating how safety mechanisms can be systematically removed.

Loading preview...