INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

dam

-0.66

UGH

-0.65

ļ

-0.64

 retarded

-0.63

 hooks

-0.60

 dick

-0.59

otrop

-0.57

 Morse

-0.56

 recount

-0.56

 probation

-0.55

POSITIVE LOGITS

asin

0.71

perse

0.71

aturdays

0.66

phas

0.65

ichick

0.64

hend

0.63

hov

0.63

unia

0.62

Shop

0.59

seys

0.59

Activations Density 0.000%

No Known Activations

This feature has no known activations.