INDEX

Explanations

news headlines or article sections marked as "RELATED" or "READ MORE"

oai_token-act-pair · gpt-3.5-turbo

references to related content or articles

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

raf

-0.72

rays

-0.70

angers

-0.66

stood

-0.63

phia

-0.63

heit

-0.61

bats

-0.59

ongs

-0.59

 lucky

-0.59

 luck

-0.58

POSITIVE LOGITS

 STOR

1.38

 STORY

1.29

 ALSO

1.27

 REPORT

1.27

WATCH

1.26

 BELOW

1.25

 VIDEOS

1.25

 VIDEO

1.25

 INFORMATION

1.22

 IMAGES

1.16

Activations Density 0.041%

news headlines or article sections marked as "RELATED" or "READ MORE"

references to related content or articles

No Comments

No Known Activations