INDEX

Explanations

the word "passengers"

oai_token-act-pair · gpt-3.5-turbo

mentions of passengers

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

arios

-0.73

 Skydragon

-0.67

SPONSORED

-0.66

enegger

-0.65

ector

-0.62

bis

-0.61

OLOGY

-0.60

laus

-0.60

 Nature

-0.60

ARA

-0.59

POSITIVE LOGITS

 passengers

0.99

hip

0.89

 passenger

0.83

 fares

0.82

 onboard

0.80

engers

0.79

iences

0.78

cript

0.75

 aboard

0.75

wings

0.75

Activations Density 0.013%

the word "passengers"

mentions of passengers

No Comments

No Known Activations