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

omination

-0.71

ovich

-0.68

omin

-0.66

 chopping

-0.66

ourney

-0.66

olit

-0.62

oking

-0.60

Compar

-0.59

iance

-0.58

 pacing

-0.58

POSITIVE LOGITS

methyl

0.75

printed

0.68

 Roche

0.65

 Niet

0.65

bern

0.64

Catal

0.64

Obj

0.64

draw

0.63

Ges

0.63

irlf

0.63

Activations Density 0.000%

No Known Activations

This feature has no known activations.