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

âĢ

-0.70

ARM

-0.69

ISS

-0.68

Space

-0.67

"],"

-0.65

Nom

-0.64

Enlarge

-0.64

cest

-0.63

":"

-0.62

âĺħâĺħ

-0.62

POSITIVE LOGITS

abo

0.72

uts

0.70

 continuity

0.68

hern

0.68

 ãĤµãĥ¼ãĥĨãĤ£ãĥ¯ãĥ³

0.67

swick

0.65

utenberg

0.65

zek

0.65

iosyn

0.65

 looph

0.65

Activations Density 0.000%

No Known Activations

This feature has no known activations.