INDEX

Explanations

statistics and information related to health issues and conditions, particularly in the United States

oai_token-act-pair · gpt-3.5-turbo Triggered by @bot

phrases related to health and mortality statistics

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 12-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.hook_resid_post

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.11.hook_resid_post

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

enegger

-0.83

eele

-0.81

swer

-0.79

£ı

-0.72

ancial

-0.68

ÃĽÃĽ

-0.67

DragonMagazine

-0.66

unity

-0.65

ERSON

-0.65

ļéĨĴ

-0.65

POSITIVE LOGITS

 lowest

1.33

 highest

1.24

 smallest

1.20

 widest

1.16

 worst

1.13

 largest

1.13

 fastest

1.12

 poorest

1.11

 busiest

1.09

 weakest

1.09

Activations Density 0.148%

statistics and information related to health issues and conditions, particularly in the United States

phrases related to health and mortality statistics

No Comments

No Known Activations