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

grave

-0.79

ero

-0.66

imental

-0.64

eria

-0.64

uvian

-0.63

Policy

-0.63

 ment

-0.62

REC

-0.61

owan

-0.60

hold

-0.59

POSITIVE LOGITS

@#&

0.91

 confir

0.79

etheless

0.73

ulhu

0.68

£ı

0.68

 heart

0.65

 conclusion

0.63

iffs

0.62

 least

0.62

 referen

0.61

Activations Density 0.000%

No Known Activations

This feature has no known activations.