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

 jelly

-0.67

ijing

-0.66

 Jonah

-0.65

adan

-0.65

angan

-0.64

Ã£

-0.62

ibr

-0.62

erm

-0.62

ghan

-0.62

icion

-0.62

POSITIVE LOGITS

enstein

0.73

lihood

0.72

rawdownloadcloneembedreportprint

0.71

Benz

0.70

BuyableInstoreAndOnline

0.70

 "$:/

0.70

FTWARE

0.69

Keefe

0.69

redits

0.68

Cooldown

0.67

Activations Density 0.000%

No Known Activations

This feature has no known activations.