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

aho

-0.80

venth

-0.78

omi

-0.70

oha

-0.69

waukee

-0.65

«ĺ

-0.65

obi

-0.65

omsday

-0.64

osexual

-0.64

âĸĪâĸĪâĸĪâĸĪâĸĪâĸĪâĸĪâĸĪ

-0.63

POSITIVE LOGITS

uits

0.78

auld

0.68

 Planet

0.66

Kut

0.65

 Niet

0.65

iffe

0.65

kers

0.64

 McDonnell

0.64

 ---------

0.62

 overd

0.62

Activations Density 0.000%

No Known Activations

This feature has no known activations.