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

externalActionCode

-0.69

eret

-0.66

vP

-0.65

ople

-0.65

 Sessions

-0.64

 Vanderbilt

-0.64

\'

-0.63

met

-0.62

put

-0.62

umph

-0.62

POSITIVE LOGITS

 board

2.63

Board

1.84

 Board

1.56

 boards

1.52

board

1.47

 Boards

1.22

boards

1.08

boarding

0.80

 boarded

0.76

eering

0.76

Activations Density 0.000%

No Known Activations

This feature has no known activations.