INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

VD

-0.81

KEN

-0.73

Cyp

-0.71

DEV

-0.69

ARB

-0.68

Lag

-0.66

COL

-0.66

Zip

-0.66

 Gorge

-0.66

DOWN

-0.65

POSITIVE LOGITS

ami

0.78

meal

0.78

immigrant

0.70

kowski

0.67

itial

0.65

anish

0.65

riots

0.65

ichick

0.64

 apiece

0.64

osi

0.63

Activations Density 0.000%

No Known Activations

This feature has no known activations.