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

olicy

-0.83

elist

-0.72

ENA

-0.69

rained

-0.68

inant

-0.67

Reply

-0.65

)=(

-0.65

olin

-0.65

SPONSORED

-0.64

ificant

-0.62

POSITIVE LOGITS

phe

0.71

quickShipAvailable

0.65

washer

0.65

 squared

0.65

Sod

0.63

raltar

0.61

hower

0.61

eger

0.60

 Beau

0.59

reet

0.59

Activations Density 0.000%

No Known Activations

This feature has no known activations.