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

iple

-0.74

doors

-0.71

endum

-0.71

emale

-0.69

emonium

-0.69

roleum

-0.69

addons

-0.66

ujah

-0.66

after

-0.65

leased

-0.65

POSITIVE LOGITS

 Hastings

0.72

 Hockey

0.68

estern

0.67

 Stevenson

0.65

 Simulator

0.64

 Credit

0.63

 grav

0.63

 Sparkle

0.62

hirt

0.62

 Rooney

0.62

Activations Density 0.000%

No Known Activations

This feature has no known activations.