INDEX

Explanations

information related to health, specifically focusing on potential risks, diseases, and warnings

oai_token-act-pair · gpt-3.5-turbo

terms and phrases associated with health issues and medical conditions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

 Helpful

-0.84

coni

-0.81

 Reloaded

-0.77

 Hundred

-0.73

ãĥĺ

-0.66

 Guilty

-0.66

Ń·

-0.64

âĸ¬

-0.64

 Pixel

-0.63

 Darkness

-0.62

POSITIVE LOGITS

 care

1.08

care

1.03

iest

1.02

span

1.01

ily

0.97

fulness

0.93

aceutical

0.90

 professionals

0.89

 isot

0.87

iness

0.85

Activations Density 0.030%

information related to health, specifically focusing on potential risks, diseases, and warnings

terms and phrases associated with health issues and medical conditions

No Comments

No Known Activations