INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

Hart

-0.74

Se

-0.68

ng

-0.67

Rain

-0.67

Tang

-0.65

Rum

-0.65

Ser

-0.65

 Tian

-0.64

Represent

-0.64

Rect

-0.63

POSITIVE LOGITS

reditary

0.76

acca

0.76

utm

0.73

merce

0.70

anchester

0.70

cknow

0.70

ĪĴ

0.70

quickShipAvailable

0.69

ufact

0.68

nikov

0.67

Activations Density 0.000%

No Known Activations

This feature has no known activations.