INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

poke

-0.79

ãĤĵ

-0.73

 Tsukuyomi

-0.68

sth

-0.68

 Bundy

-0.65

Ark

-0.64

 SetFontSize

-0.64

ÙĴ

-0.63

cellence

-0.63

hof

-0.62

POSITIVE LOGITS

 contrace

0.83

Downloadha

0.79

 therap

0.77

brow

0.72

 conflic

0.69

 constitu

0.66

 Palestin

0.64

lean

0.63

 endorsements

0.63

BI

0.63

Activations Density 0.000%

No Known Activations

This feature has no known activations.