INDEX

Explanations

words related to computer programming concepts and activities

oai_token-act-pair · gpt-3.5-turbo

references to parsed code and technical elements

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.8.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 seiz

-0.85

 dexter

-0.76

tal

-0.75

fid

-0.74

 circumst

-0.74

 polic

-0.73

fig

-0.73

 ivory

-0.72

 sacrific

-0.71

 epid

-0.71

POSITIVE LOGITS

World

1.37

Burn

1.36

Bl

1.30

Win

1.29

Pain

1.27

Party

1.27

Talk

1.27

Cut

1.26

Program

1.26

Current

1.26

Activations Density 0.117%

words related to computer programming concepts and activities

references to parsed code and technical elements

No Comments

No Known Activations