INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

 Administ

-0.81

 ç¥ŀ

-0.80

ificant

-0.76

terday

-0.74

ãĥĳ

-0.73

 Lumpur

-0.71

âķĲâķĲ

-0.71

theless

-0.70

itaire

-0.68

etimes

-0.65

POSITIVE LOGITS

onto

0.72

 Pearl

0.68

amo

0.67

oney

0.66

Smith

0.65

icals

0.64

ofi

0.64

 Laurel

0.64

rail

0.63

UCS

0.63

Activations Density 0.000%

No Known Activations

This feature has no known activations.