INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

irlf

-0.81

Adams

-0.75

DEC

-0.75

MpServer

-0.72

println

-0.71

jet

-0.68

ascus

-0.67

inances

-0.67

accompan

-0.66

erman

-0.66

POSITIVE LOGITS

 captcha

0.79

ãĥ³ãĤ¸

0.75

 chase

0.75

 Percent

0.73

 Champions

0.69

 halftime

0.68

 Hood

0.67

 lane

0.66

 dermat

0.63

tag

0.63

Activations Density 0.000%

No Known Activations

This feature has no known activations.