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

 Taken

-0.63

ocide

-0.57

Īè

-0.56

emon

-0.56

itis

-0.55

icut

-0.54

rider

-0.54

iste

-0.54

orney

-0.53

lesh

-0.53

POSITIVE LOGITS

 themselves

1.84

their

1.67

 their

1.54

they

1.41

 they

1.37

 theirs

1.32

Their

1.31

 THEIR

1.29

They

1.18

 Their

1.15

Activations Density 0.000%

No Known Activations

This feature has no known activations.