INDEX

Explanations

instances where the author is trying to do something

oai_token-act-pair · gpt-3.5-turbo

phrases that express attempts or efforts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

 Returning

-0.70

ELD

-0.67

 Printed

-0.63

Flo

-0.63

ILY

-0.63

 Removal

-0.62

 Lands

-0.60

ITIES

-0.59

many

-0.59

 Clouds

-0.58

POSITIVE LOGITS

 emulate

1.33

 imitate

1.29

 convince

1.24

 recreate

1.24

 replicate

1.22

 minimize

1.15

 persuade

1.14

 revive

1.12

 reconcile

1.09

 maximize

1.08

Activations Density 0.106%

instances where the author is trying to do something

phrases that express attempts or efforts

No Comments

No Known Activations