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

mere

-0.77

LET

-0.68

heter

-0.66

Brother

-0.66

mate

-0.65

EngineDebug

-0.64

senal

-0.63

AMY

-0.63

hett

-0.63

 unmarked

-0.63

POSITIVE LOGITS

awar

0.73

ansson

0.72

@#&

0.71

avascript

0.66

asma

0.65

ogical

0.63

ophysical

0.62

ippi

0.60

 Huss

0.60

 answ

0.60

Activations Density 0.000%

No Known Activations

This feature has no known activations.