INDEX

Explanations

words related to medical conditions and organizations such as "Disease," "CDC," and "epidemiological."

oai_token-act-pair · gpt-3.5-turbo

terms related to diseases and public health organizations

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

dyl

-0.68

edom

-0.67

 fracking

-0.66

jug

-0.66

vert

-0.65

ĳ

-0.65

KEN

-0.65

 chrom

-0.64

chel

-0.63

 unpaid

-0.63

POSITIVE LOGITS

 Disease

1.35

 Outbreak

1.25

 Infect

1.01

 Diseases

1.01

 Rating

0.93

 outbreak

0.89

 Massacre

0.89

 Disorders

0.88

 Syndrome

0.87

 Virus

0.86

Activations Density 0.017%

words related to medical conditions and organizations such as "Disease," "CDC," and "epidemiological."

terms related to diseases and public health organizations

No Comments

No Known Activations