INDEX

Explanations

phrases used for emphasizing innovation, importance, choices, and system improvements

oai_token-act-pair · gpt-3.5-turbo

phrases indicating actions or events that are impactful or significant

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

etheless

-0.83

Finally

-0.69

anything

-0.66

 nonetheless

-0.65

iren

-0.64

pei

-0.64

yip

-0.63

EStream

-0.63

rentices

-0.61

patrick

-0.61

POSITIVE LOGITS

 tremendous

0.68

 inaccurate

0.66

 great

0.66

omical

0.65

 untrue

0.62

 gorgeous

0.61

 immense

0.60

ometric

0.59

 visually

0.58

 aest

0.58

Activations Density 0.081%

phrases used for emphasizing innovation, importance, choices, and system improvements

phrases indicating actions or events that are impactful or significant

No Comments

No Known Activations