INDEX

Explanations

instances of the word "promoted"

oai_token-act-pair · gpt-3.5-turbo

terms associated with the concept of promotion or advancement

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

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

igh

-0.74

mount

-0.69

Fug

-0.68

kn

-0.65

 Sens

-0.64

iture

-0.63

ivan

-0.63

iddle

-0.62

knife

-0.61

Ã°

-0.61

POSITIVE LOGITS

 promoted

3.66

 promotion

2.21

 promotions

1.75

 promotes

1.64

 Promotion

1.63

 promot

1.63

 promote

1.62

 touted

1.57

 championed

1.56

 advertised

1.51

Activations Density 0.014%

instances of the word "promoted"

terms associated with the concept of promotion or advancement

No Comments

No Known Activations