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

urdue

-0.74

"@

-0.72

ptive

-0.71

pal

-0.70

Ã±

-0.70

Pau

-0.69

edia

-0.67

illac

-0.65

chev

-0.65

MPG

-0.63

POSITIVE LOGITS

elvet

0.74

Revolution

0.69

 chancellor

0.67

Reviewed

0.66

=-=-=-=-=-=-=-=-

0.66

akening

0.65

lihood

0.63

acea

0.62

 wrought

0.62

Disclaimer

0.61

Activations Density 0.000%

No Known Activations

This feature has no known activations.