INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

atoon

-0.78

ategory

-0.77

seless

-0.73

porary

-0.73

tein

-0.72

iction

-0.71

emale

-0.70

icted

-0.69

 showers

-0.68

ricanes

-0.68

POSITIVE LOGITS

ci

0.73

mother

0.69

father

0.68

 neighb

0.65

sen

0.65

gemony

0.65

 influ

0.63

 Palest

0.63

Eth

0.63

 Ariel

0.63

Activations Density 0.000%

No Known Activations

This feature has no known activations.