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

çļ

-0.76

æł

-0.73

URI

-0.71

©

-0.69

chairs

-0.69

ļéĨĴ

-0.68

 chairs

-0.68

 exha

-0.67

âĦ¢:

-0.67

 exting

-0.66

POSITIVE LOGITS

Stock

0.68

GM

0.66

stock

0.65

GN

0.64

 Sabb

0.64

 slaying

0.63

stead

0.62

vari

0.61

urgy

0.60

ospace

0.60

Activations Density 0.000%

No Known Activations

This feature has no known activations.