INDEX

Explanations

commands or instructions

oai_token-act-pair · gpt-3.5-turbo

references to leadership or authority figures

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

 condem

-0.73

 passions

-0.67

 conduc

-0.66

 clitor

-0.66

 vulner

-0.66

 flowering

-0.66

 pleasures

-0.66

naires

-0.66

 phosphate

-0.65

 itiner

-0.63

POSITIVE LOGITS

âĶĢâĶĢ

1.13

ï¸ı

1.11

âĶĢâĶĢâĶĢâĶĢ

0.88

âķĲâķĲ

0.88

ICE

0.87

ever

0.83

âĶĢâĶĢâĶĢâĶĢâĶĢâĶĢâĶĢâĶĢ

0.81

conom

0.80

athom

0.79

Â§

0.78

Activations Density 0.222%

commands or instructions

references to leadership or authority figures

No Comments

No Known Activations