INDEX

Explanations

quotes or pieces of advice

oai_token-act-pair · gpt-3.5-turbo

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

agra

-0.71

ERY

-0.70

 effected

-0.67

alysed

-0.66

quin

-0.64

azo

-0.63

atorial

-0.61

agram

-0.60

rift

-0.60

retched

-0.59

POSITIVE LOGITS

 beware

1.27

Always

1.26

 Avoid

1.22

don

1.21

DON

1.18

 ALWAYS

1.17

 Always

1.16

 NEVER

1.16

Never

1.16

 avoid

1.15

Activations Density 0.356%

quotes or pieces of advice

No Comments

No Known Activations