llama-3-70B-Instruct-abliterated
This is meta-llama/Llama-3-70B-Instruct with orthogonalized bfloat16 safetensor weights, generated with the methodology that was described in the preview paper/blog post: 'Refusal in LLMs is mediated by a single direction' which I encourage you to read to understand more.
text → text · made by failspy
Sold by 1 way
| Seller | Lane | Rate |
|---|---|---|
| Featherless AI aggregator | standard | $3.5per Mtok in$8.3per Mtok outapi.featherless.ai · read 2026-09-20 |
About
llama-3-70B-Instruct-abliterated — a text model from failspy, sold by one company from $3.5 in and $8.3 out per million tokens.
It takes text and returns text, with a context window of 8,192 tokens. It was published in May 2024. The catalogue files it under chat. Its weights are published under the llama3 licence, which attaches conditions the plain open licences do not. Only Featherless AI sells it, at $3.5 in and $8.3 out per million tokens.
Every current figure
- Maker
- failspy
- Register
- model
- Takes
- text
- Returns
- text
- Context
- 8,192 tokens
- Published
- May 2024
- Parameters
- 70 billion · read from its own name
- Licence
- llama3
- Sellers
- 1
- Maker's own price
- not read
- Price
- $3.5 in and $8.3 out per million tokens — Featherless AI
Known as 4 names
failspy/llama-3-70B-Instruct-abliteratedGreytechai/Llama-3-70B-Instruct-abliterated-v3failspy/Llama-3-70B-Instruct-abliterated-v3nguyenthilaitrieulong/llama-3-70B-Instruct-abliterated