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

icone

-0.78

uca

-0.73

onen

-0.71

ãħĭ

-0.68

kefeller

-0.67

NYSE

-0.66

 Scotia

-0.65

unic

-0.65

uin

-0.63

 Omega

-0.63

POSITIVE LOGITS

sect

0.78

arr

0.76

arre

0.70

duction

0.68

amer

0.68

ilit

0.67

omes

0.67

ames

0.66

cknowled

0.64

GER

0.63

Activations Density 0.000%

No Known Activations

This feature has no known activations.