INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

DonaldTrump

-0.82

ially

-0.79

elines

-0.78

ħĭ

-0.74

angles

-0.74

ographically

-0.74

adelphia

-0.74

union

-0.73

argo

-0.73

eline

-0.72

POSITIVE LOGITS

 Hare

0.78

Fah

0.76

 Faul

0.68

 alle

0.68

 Ivory

0.66

 laps

0.65

 Guides

0.65

Hum

0.64

Pag

0.64

Sph

0.63

Activations Density 0.000%

No Known Activations

This feature has no known activations.