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

edi

-0.73

assies

-0.70

ussed

-0.70

uss

-0.64

ppa

-0.64

riers

-0.64

ingu

-0.64

spo

-0.62

 Whilst

-0.62

ulla

-0.62

POSITIVE LOGITS

 Michaels

0.66

 Beir

0.65

 Akron

0.65

 Dare

0.65

oret

0.64

 grandparents

0.63

Downloadha

0.60

tones

0.58

 Malk

0.57

 Norn

0.57

Activations Density 0.000%

No Known Activations

This feature has no known activations.