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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

mast

-0.71

erous

-0.68

rpm

-0.65

tick

-0.64

etr

-0.64

ãĤ©

-0.64

 Lowe

-0.63

frey

-0.63

 Tick

-0.62

 Hung

-0.62

POSITIVE LOGITS

engers

0.72

ieu

0.62

 proxies

0.62

 advoc

0.61

riages

0.58

 Cheong

0.57

 behaviours

0.57

esters

0.57

uscript

0.57

 screenshots

0.56

Activations Density 0.000%

No Known Activations

This feature has no known activations.