INDEX

Explanations

statistics or numerical data related to particular categories such as animals, government, or medical conditions

oai_token-act-pair · gpt-3.5-turbo Triggered by @bot

phrases indicating the source or origin of various issues or concerns

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 12-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.hook_resid_post

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.11.hook_resid_post

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

raq

-0.92

ynes

-0.88

miah

-0.82

framework

-0.81

ppa

-0.80

arity

-0.78

mil

-0.77

arse

-0.76

ftime

-0.76

unny

-0.75

POSITIVE LOGITS

 harm

1.09

 ridicule

1.09

 exploitation

1.09

 predators

1.08

 prosecution

1.05

 attack

1.03

 scrutiny

1.00

 predatory

1.00

 liability

0.99

 extinction

0.99

Activations Density 0.163%

statistics or numerical data related to particular categories such as animals, government, or medical conditions

phrases indicating the source or origin of various issues or concerns

No Comments

No Known Activations