INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

Reviewed

-0.68

Thumbnail

-0.66

ĪĴ

-0.66

 comma

-0.65

 euth

-0.63

 filled

-0.63

 rejo

-0.63

ĻĤ

-0.63

 exercised

-0.62

Ħ¢

-0.62

POSITIVE LOGITS

 VIDEOS

0.71

adr

0.68

anz

0.66

 DRAG

0.63

zos

0.63

sche

0.63

ures

0.62

erences

0.62

emet

0.61

eny

0.61

Activations Density 0.000%

No Known Activations

This feature has no known activations.