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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

morrow

-0.72

rencies

-0.72

ebook

-0.71

lip

-0.69

 travellers

-0.66

cgi

-0.64

=-=-=-=-=-=-=-=-

-0.64

 raids

-0.63

awei

-0.62

owship

-0.62

POSITIVE LOGITS

 Sale

0.70

 Sell

0.68

ment

0.66

Tad

0.65

 Share

0.65

Dah

0.64

Sal

0.64

raised

0.64

ained

0.63

ured

0.63

Activations Density 0.000%

No Known Activations

This feature has no known activations.