INDEX

Explanations

phrases prompting action or investigation

oai_token-act-pair · gpt-3.5-turbo

phrases that initiate a suggestion or invitation to do something

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

stadt

-0.64

mouth

-0.59

ads

-0.58

 pockets

-0.58

 coded

-0.58

assian

-0.57

ious

-0.57

 invasive

-0.56

yr

-0.56

 carc

-0.55

POSITIVE LOGITS

Let

3.16

Let

2.34

 Lets

2.15

LET

1.73

let

1.70

let

1.40

 Suppose

1.38

 Forget

1.38

LET

1.32

 Take

1.28

Activations Density 0.012%

phrases prompting action or investigation

phrases that initiate a suggestion or invitation to do something

No Comments

No Known Activations