INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

 captive

-0.70

jar

-0.67

 throat

-0.66

ple

-0.65

atio

-0.64

 mort

-0.62

 disple

-0.62

ching

-0.62

quartered

-0.61

 heir

-0.60

POSITIVE LOGITS

uggest

0.74

IVERS

0.68

rences

0.68

glomer

0.66

allas

0.65

baum

0.65

 antid

0.63

 Wildcats

0.63

olkien

0.63

yrinth

0.63

Activations Density 0.000%

No Known Activations

This feature has no known activations.