INDEX

Explanations

warnings about graphic or explicit content along with mentions of language that may not be suitable for all audiences

oai_token-act-pair · gpt-3.5-turbo

warnings or notifications about graphic content

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

 anew

-0.83

sonian

-0.81

externalActionCode

-0.73

 contempor

-0.71

Newsletter

-0.70

Reviewer

-0.69

united

-0.66

DragonMagazine

-0.66

 Gutenberg

-0.66

 reinvest

-0.65

POSITIVE LOGITS

 dangers

1.01

danger

0.99

 beware

0.97

 dangerous

0.93

 danger

0.92

 abuses

0.87

 foul

0.87

 scares

0.84

 injure

0.84

 unsafe

0.84

Activations Density 0.635%

warnings about graphic or explicit content along with mentions of language that may not be suitable for all audiences

warnings or notifications about graphic content

No Comments

No Known Activations