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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

nered

-0.66

 curriculum

-0.64

spons

-0.63

recy

-0.61

irth

-0.60

advert

-0.58

 secretive

-0.57

ipation

-0.57

slow

-0.57

 inertia

-0.57

POSITIVE LOGITS

DragonMagazine

0.70

ENA

0.69

emen

0.68

ORPG

0.68

ESE

0.66

 è£ıç

0.66

acter

0.66

)=(

0.65

ãĤ¦ãĤ¹

0.64

TY

0.64

Activations Density 0.000%

No Known Activations

This feature has no known activations.