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

peria

-0.72

nces

-0.69

 honoured

-0.69

 alliances

-0.67

 levers

-0.66

sag

-0.64

plates

-0.64

 Dice

-0.64

ipped

-0.63

 tactical

-0.63

POSITIVE LOGITS

akery

0.68

atech

0.66

ership

0.64

 Ferry

0.64

assetsadobe

0.63

 Barn

0.63

ð

0.63

GM

0.63

fa

0.62

uthor

0.62

Activations Density 0.000%

No Known Activations

This feature has no known activations.