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

ascript

-0.85

conservancy

-0.73

itory

-0.72

IU

-0.71

atl

-0.71

ainer

-0.71

ickr

-0.70

itten

-0.70

cape

-0.69

illard

-0.68

POSITIVE LOGITS

 country

1.12

country

0.82

ãĥĸ

0.78

 Country

0.76

å§«

0.76

 denomination

0.72

 nation

0.70

 Countries

0.70

ãĥĻ

0.69

Country

0.69

Activations Density 0.000%

No Known Activations

This feature has no known activations.