INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

independent

-0.65

core

-0.61

atory

-0.61

cloth

-0.60

uggle

-0.59

archive

-0.58

 conduc

-0.58

chi

-0.58

 tested

-0.56

metry

-0.56

POSITIVE LOGITS

âĢ¢âĢ¢

0.74

 Bans

0.69

ihar

0.67

Tel

0.65

Cola

0.63

Lap

0.63

 Kard

0.63

 Tire

0.62

itÃ©

0.62

appa

0.62

Activations Density 0.000%

No Known Activations

This feature has no known activations.