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

èª

-0.73

iaz

-0.72

 TAMADRA

-0.69

 jails

-0.65

jong

-0.64

dos

-0.64

âĨĳ

-0.63

omics

-0.63

merce

-0.62

ki

-0.62

POSITIVE LOGITS

oun

0.71

enance

0.71

 Flavoring

0.69

erville

0.65

utonium

0.64

our

0.64

utenberg

0.63

 Norn

0.62

pers

0.61

 Polo

0.61

Activations Density 0.000%

No Known Activations

This feature has no known activations.