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

ĸļ

-1.03

ãĥķãĤ©

-0.81

leeve

-0.79

AppData

-0.79

ãĤ¼ãĤ¦ãĤ¹

-0.78

ãĤ¦ãĤ¹

-0.77

ãĥ¯ãĥ³

-0.76

 Izan

-0.71

addons

-0.69

è¦ļéĨĴ

-0.69

POSITIVE LOGITS

gression

0.76

Reviewer

0.75

hander

0.68

istry

0.67

 holiday

0.66

 neoc

0.63

 antis

0.63

ulative

0.62

asking

0.62

puting

0.62

Activations Density 0.000%

No Known Activations

This feature has no known activations.