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

DragonMagazine

-0.96

-0.80

EStream

-0.72

interstitial

-0.68

 scient

-0.63

PsyNetMessage

-0.61

 halluc

-0.61

 steroids

-0.61

UFC

-0.61

EStreamFrame

-0.60

POSITIVE LOGITS

anium

0.90

bsp

0.86

cedes

0.75

ackets

0.75

iny

0.74

acks

0.72

uka

0.72

imi

0.72

anc

0.70

sts

0.69

Activations Density 0.000%

No Known Activations

This feature has no known activations.