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

hift

-0.98

cffff

-0.84

fty

-0.77

quet

-0.76

unin

-0.74

 toile

-0.68

utan

-0.67

 antioxid

-0.66

 wrestle

-0.65

eful

-0.65

POSITIVE LOGITS

wolves

0.73

ground

0.70

IST

0.66

Aust

0.66

sburg

0.63

 centers

0.62

marks

0.60

ays

0.60

ses

0.60

iss

0.60

Activations Density 0.000%

No Known Activations

This feature has no known activations.