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

aly

-0.80

iets

-0.74

alkyrie

-0.65

 Karma

-0.65

kai

-0.65

ervatives

-0.64

ionics

-0.64

ilogy

-0.63

 toolbar

-0.62

iki

-0.61

POSITIVE LOGITS

nom

1.63

OW

0.76

NAS

0.71

vik

0.70

QU

0.69

ENC

0.68

Scotland

0.67

 Scand

0.66

 Senegal

0.65

GEN

0.65

Activations Density 0.000%

No Known Activations

This feature has no known activations.