INDEX

Explanations

references to medical patients

oai_token-act-pair · gpt-3.5-turbo

references to patients in a medical or healthcare context

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

 Languages

-0.67

athlon

-0.66

 --------------------

-0.65

 Rocks

-0.64

Lag

-0.64

Dag

-0.63

 Rings

-0.62

 Tags

-0.62

tops

-0.62

 Bows

-0.61

POSITIVE LOGITS

 patient

3.70

 Patient

2.70

patient

2.49

 patients

2.27

 Patients

2.22

 clinical

1.72

 physician

1.64

 surgeon

1.50

 nurse

1.48

 inmate

1.47

Activations Density 0.013%

references to medical patients

references to patients in a medical or healthcare context

No Comments

No Known Activations