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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ibur

-0.79

endish

-0.66

icularly

-0.63

igmatic

-0.63

rir

-0.62

dry

-0.62

 Adren

-0.62

 Gleaming

-0.61

helm

-0.61

untled

-0.61

POSITIVE LOGITS

FontSize

0.69

 shortcuts

0.60

zos

0.60

 imports

0.59

geries

0.59

etheless

0.59

auga

0.59

 Germans

0.59

gets

0.56

 alloc

0.56

Activations Density 0.000%

No Known Activations

This feature has no known activations.