Llama-3.1-8B-Instruct-heretic

Model overview

Llama-3.1-8B-Instruct-heretic is published by notcvnt as a text-generation model. The captured configuration identifies LlamaForCausalLM with model type llama, and Safetensors metadata reports 8,030,261,248 parameters. According to the model card, it is a Heretic-based decensored edit of meta-llama/Llama-3.1-8B-Instruct, and the recorded license is llama3.1.

Recorded capabilities

Use cases in the source record

• Multilingual assistant-like chat in the card's supported languages, served through the snapshot-listed conversational path subject to refresh.
• Transformers tool-calling experiments that follow the card's documented chat-template, tool-call, and tool-role message pattern.

Limitations and unknowns

• No Ethen-verified evaluation results were extracted; refusal and KL figures are publisher-reported Heretic metrics.
• Provider state is historical snapshot data and should be refreshed before being presented as current.
• Context, token, hardware, and emissions figures reproduce Meta's Llama 3.1 family description rather than a measurement of this Heretic edit.

Source and provenance

Source: notcvnt/Llama-3.1-8B-Instruct-heretic

Captured: Unknown. Processed: 2026-09-07T19:35:59.262085+00:00.

This is a decensored version of meta-llama/Llama-3.1-8B-Instruct , made using Heretic Performance Metric This model Original model KL divergence 0.042 0 Refusals 4/100 96/100 Model Information The Meta Llama 3.1 collection of multilingual large language models (LLMs) is a collection of pretrained and instruction tuned generative models in 8B, 70B and 405B sizes (text in/text out). The Llama 3.1 instruction tuned text only models (8B, 70B, 405B) are optimized for multilingual dialogue use cases and outperform many of the available open source and closed chat models on common industry benchmarks. Model developer : Meta Model Architect…