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

inational

-0.81

obi

-0.76

elf

-0.75

fulness

-0.75

irin

-0.75

Ã¸

-0.73

antic

-0.71

ourge

-0.71

cill

-0.70

agonist

-0.70

POSITIVE LOGITS

 Scrib

0.80

Magikarp

0.72

 Starting

0.69

Tee

0.68

 Scholarship

0.66

 Topic

0.66

ASA

0.65

 Dane

0.63

 Jere

0.63

 COVER

0.63

Activations Density 0.000%

No Known Activations

This feature has no known activations.