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

MAS

-0.69

 Hayward

-0.66

ĳå£«

-0.66

 Howe

-0.65

hers

-0.64

elson

-0.62

natureconservancy

-0.62

 Chavez

-0.62

ãĤ·ãĥ£

-0.61

 Greenberg

-0.60

POSITIVE LOGITS

ions

0.74

uted

0.71

arching

0.70

knit

0.68

orks

0.67

 unden

0.65

ound

0.65

 Brands

0.64

pop

0.64

ioned

0.63

Activations Density 0.000%

No Known Activations

This feature has no known activations.