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

uterte

-0.82

Bei

-0.81

hammad

-0.75

 Ezek

-0.74

odo

-0.73

omsky

-0.73

elaide

-0.72

puting

-0.71

ertodd

-0.71

enda

-0.71

POSITIVE LOGITS

isons

0.74

 Reincarn

0.69

OTT

0.66

 Psychiat

0.66

oji

0.64

 Addiction

0.63

 Mental

0.62

Maj

0.61

 Enhanced

0.61

LSD

0.59

Activations Density 0.000%

No Known Activations

This feature has no known activations.