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

 Portug

-0.71

vag

-0.68

 boarded

-0.66

iku

-0.65

aciously

-0.65

 vill

-0.64

bang

-0.63

 resc

-0.63

agle

-0.62

glers

-0.62

POSITIVE LOGITS

 Flavoring

1.12

 Whereas

0.95

 Accordingly

0.95

 Needless

0.91

 Alternatively

0.90

 Furthermore

0.90

 Notably

0.88

 Additionally

0.87

 Consequently

0.87

 Moreover

0.86

Activations Density 0.000%

No Known Activations

This feature has no known activations.