INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

stakes

-0.76

£ı

-0.68

teen

-0.66

 arenas

-0.66

nation

-0.65

wagen

-0.65

Te

-0.65

bats

-0.64

 nonex

-0.63

worldly

-0.63

POSITIVE LOGITS

aucus

0.72

cific

0.70

UGE

0.68

issors

0.67

urus

0.66

arettes

0.66

isse

0.64

 consolidated

0.64

mus

0.62

rique

0.62

Activations Density 0.000%

No Known Activations

This feature has no known activations.