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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ratulations

-0.73

bably

-0.72

biz

-0.71

fit

-0.68

vre

-0.68

profile

-0.65

fitting

-0.64

fits

-0.63

alore

-0.61

recent

-0.59

POSITIVE LOGITS

Cra

0.69

Creat

0.66

CSS

0.64

imeter

0.64

ãģ®å®

0.62

 Fathers

0.62

HRC

0.62

Ire

0.61

Cro

0.60

assetsadobe

0.59

Activations Density 0.000%

No Known Activations

This feature has no known activations.