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

rawdownloadcloneembedreportprint

-0.78

 mort

-0.68

illon

-0.61

 incompet

-0.61

 sprink

-0.60

 Flavoring

-0.59

 intimid

-0.59

Brit

-0.59

 srfAttach

-0.58

£ı

-0.58

POSITIVE LOGITS

opolis

0.74

wcs

0.69

 Hunters

0.65

ourses

0.65

 Jericho

0.64

 Boys

0.64

sers

0.64

oter

0.63

agara

0.63

 Owens

0.63

Activations Density 0.000%

No Known Activations

This feature has no known activations.