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

Ô

-0.88

license

-0.76

 Flavoring

-0.67

nit

-0.67

edia

-0.66

otrop

-0.64

Flag

-0.64

arine

-0.63

Maps

-0.63

Art

-0.62

POSITIVE LOGITS

shall

0.71

will

0.69

arious

0.68

 will

0.68

 doom

0.62

luent

0.61

aukee

0.61

 Pilgrim

0.61

 attendant

0.61

aucus

0.60

Activations Density 0.000%

No Known Activations

This feature has no known activations.