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

tu

-0.67

iesel

-0.66

soDeliveryDate

-0.65

 instructor

-0.65

igs

-0.64

eno

-0.62

isSpecialOrderable

-0.60

 accuser

-0.60

 switches

-0.59

avascript

-0.59

POSITIVE LOGITS

âĢ

0.89

ãĢ

0.87

Sund

0.79

SHIP

0.79

song

0.74

Austral

0.72

ãĤ¼ãĤ¦ãĤ¹

0.67

llan

0.67

âĺħ

0.66

naire

0.65

Activations Density 0.000%

No Known Activations

This feature has no known activations.