INDEX

Explanations

medication-related words and phrases, especially those related to treating various medical conditions

oai_token-act-pair · gpt-3.5-turbo

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

DragonMagazine

-0.98

soDeliveryDate

-0.86

sbm

-0.81

displayText

-0.81

Catalog

-0.81

fman

-0.77

soType

-0.75

isSpecialOrderable

-0.75

cknowled

-0.74

çķ

-0.73

POSITIVE LOGITS

 ailments

1.58

 arthritis

1.35

 illnesses

1.34

 diseases

1.33

 inflammation

1.32

 nausea

1.29

 headaches

1.28

 cancers

1.26

 insomnia

1.25

 cancer

1.25

Activations Density 0.191%

medication-related words and phrases, especially those related to treating various medical conditions

No Comments

No Known Activations