INDEX

Explanations

question words at the beginning of interrogative sentences

oai_token-act-pair · gpt-3.5-turbo

instances of the phrase "Are you"

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

 rather

-0.73

 scrap

-0.67

 special

-0.64

 credit

-0.63

 priv

-0.63

 half

-0.63

 favored

-0.62

lim

-0.61

 printing

-0.61

 signature

-0.59

POSITIVE LOGITS

Are

3.14

Were

1.91

Have

1.80

Is

1.75

Are

1.74

Aren

1.65

Did

1.64

Do

1.61

Can

1.52

Could

1.51

Activations Density 0.014%

question words at the beginning of interrogative sentences

instances of the phrase "Are you"

No Comments

No Known Activations