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

acus

-0.75

 refuel

-0.67

 indent

-0.67

 redes

-0.67

 datas

-0.67

arthed

-0.67

 brainstorm

-0.66

 morphine

-0.65

oÄŁ

-0.64

xual

-0.63

POSITIVE LOGITS

hammer

0.91

rier

0.77

urai

0.65

rab

0.64

FF

0.62

Smith

0.62

 Smith

0.61

cat

0.61

Rag

0.60

Activations Density 0.000%

No Known Activations

This feature has no known activations.