INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

amount

-0.67

 redesign

-0.65

Ĥİ

-0.63

 institute

-0.57

 tigers

-0.56

apo

-0.56

uni

-0.56

 overest

-0.55

 appra

-0.55

 Associates

-0.55

POSITIVE LOGITS

 sorts

0.98

 course

0.86

icial

0.79

course

0.74

irable

0.70

rored

0.69

onom

0.67

 Mour

0.66

ursday

0.65

sky

0.64

Activations Density 0.000%

No Known Activations

This feature has no known activations.