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

hedon

-0.79

oms

-0.74

ody

-0.69

ecause

-0.69

olic

-0.69

icz

-0.68

anski

-0.68

odies

-0.65

yang

-0.65

matter

-0.64

POSITIVE LOGITS

cffffcc

0.81

ãĤ©

0.76

ãĤ¿

0.71

ãĥīãĥ©

0.69

ressor

0.68

ãĥŁ

0.66

ãĥĢ

0.66

ãĤ¼ãĤ¦ãĤ¹

0.66

ãĥĻ

0.64

 guiActiveUn

0.63

Activations Density 0.000%

No Known Activations

This feature has no known activations.