Blog

Discussing and disseminating open-source AI research.

2023

May

馃惗Safetensors audited as really safe and becoming the default

May 23, 2023 路 Nicolas Patry, Stella Biderman, Garry Jean-Baptiste

Alignment Research @ EleutherAI

May 3, 2023 路 Curtis Huebner

April

Transformer Math 101

April 18, 2023 路 Quentin Anthony, Stella Biderman, Hailey Schoelkopf

Exploratory Analysis of TRLX RLHF Transformers with TransformerLens

April 2, 2023 路 Curt Tigges

March

EleutherAI Second Retrospective: The long version

March 26, 2023 路 Stella Biderman, Shivanshu Purohit, Curtis Huebner, Leo Gao, Connor Leahy, Eric Hallahan

The View from 30,000 Feet: Preface to the Second EleutherAI Retrospective

March 2, 2023 路 Stella Biderman, Curtis Huebner, Connor Leahy, Eric Hallahan

2022

February

Announcing GPT-NeoX-20B

February 2, 2022 路 Connor Leahy

2021

October

A Preliminary Exploration into Factored Cognition with Language Models

October 25, 2021 路 Leo Gao, Kyle McDonell, Laria Reynolds, Stella Biderman

Multiple Choice Normalization in LM Evaluation

October 11, 2021 路 Leo Gao

August

Downstream Evaluations of Rotary Position Embeddings

August 16, 2021 路 Leo Gao

July

What A Long, Strange Trip It's Been: EleutherAI One Year Retrospective

July 7, 2021 路 Connor Leahy, Eric Hallahan, Leo Gao, Stella Biderman

June

Why Release a Large Language Model?

June 2, 2021 路 Connor Leahy

May

On the Sizes of OpenAI API Models

May 24, 2021 路 Leo Gao

Evaluating Different Fewshot Description Prompts on GPT-3

May 24, 2021 路 Leo Gao

Finetuning Models on Downstream Tasks

May 24, 2021 路 Leo Gao

Activation Function Ablation

May 24, 2021 路 Leo Gao

April

Rotary Embeddings: A Relative Revolution

April 20, 2021 路 Stella Biderman, Sid Black, Charles Foster, Leo Gao, Eric Hallahan, Horace He, Ben Wang, Phil Wang