INDEX

Explanations

phrases related to public statements or mentions

oai_token-act-pair · gpt-3.5-turbo

references to public statements or actions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.hook_resid_pre

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

nesota

-0.87

nian

-0.80

 Clever

-0.79

 Gamb

-0.74

Sap

-0.73

akings

-0.70

 Bagg

-0.70

 Salvador

-0.70

orsi

-0.69

akeru

-0.68

POSITIVE LOGITS

 publicly

0.89

 RELE

0.87

 divul

0.83

ãĤ©

0.81

 disclosed

0.81

isSpecialOrderable

0.78

 distingu

0.77

 disav

0.76

 VIDEOS

0.76

 disclosing

0.75

Activations Density 0.006%

phrases related to public statements or mentions

references to public statements or actions

No Comments

No Known Activations