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

alid

-0.82

awks

-0.78

igators

-0.70

ifa

-0.68

urities

-0.66

stones

-0.64

idable

-0.64

 Purpose

-0.64

wered

-0.63

roads

-0.63

POSITIVE LOGITS

 reluct

0.80

 hither

0.74

 veter

0.73

 mosqu

0.73

LY

0.72

bilt

0.69

 tiss

0.69

 exha

0.67

ROM

0.65

 Kass

0.64

Activations Density 0.000%

No Known Activations

This feature has no known activations.