INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

SpaceEngineers

-0.78

 Passage

-0.66

iHUD

-0.61

 Termin

-0.60

Water

-0.60

Termin

-0.59

etsk

-0.58

 Cosponsors

-0.58

 scenery

-0.57

bye

-0.56

POSITIVE LOGITS

mania

0.70

Ã§a

0.70

esson

0.69

 Bleach

0.67

tein

0.65

ruary

0.64

agall

0.63

chester

0.63

ovie

0.63

cgi

0.63

Activations Density 0.000%

No Known Activations

This feature has no known activations.