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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 ILCS

-0.92

aeda

-0.90

ategory

-0.89

avorite

-0.80

ucket

-0.78

theless

-0.76

hovah

-0.76

backer

-0.72

aphael

-0.72

aeper

-0.68

POSITIVE LOGITS

tics

0.70

...)

0.65

ber

0.65

?)

0.63

Ã©t

0.62

 discont

0.62

ovo

0.61

âĢ¦)

0.60

!)

0.60

ually

0.59

Activations Density 0.000%

No Known Activations

This feature has no known activations.