INDEX

Explanations

medical and health-related terms, including factors like medications, exercise, and dietary intake

oai_token-act-pair · gpt-3.5-turbo

biomedical and health-related terms and concepts

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

DragonMagazine

-0.83

Tokens

-0.82

ateurs

-0.77

emonium

-0.77

named

-0.77

nels

-0.76

}}}

-0.76

echo

-0.75

Grid

-0.75

csv

-0.73

POSITIVE LOGITS

 alcoholism

1.51

 marital

1.33

 obesity

1.33

 illness

1.28

 diabetes

1.28

 alcohol

1.28

 depression

1.27

 hypertension

1.25

 allergies

1.24

 malnutrition

1.22

Activations Density 0.257%

medical and health-related terms, including factors like medications, exercise, and dietary intake

biomedical and health-related terms and concepts

No Comments

No Known Activations