INDEX

Explanations

medical conditions, particularly epilepsy and asthma

oai_token-act-pair · gpt-3.5-turbo

references to neurological and respiratory medical conditions

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

andum

-0.75

eworld

-0.71

agher

-0.68

achu

-0.66

ord

-0.66

izens

-0.64

aunder

-0.64

sa

-0.63

spring

-0.62

parts

-0.62

POSITIVE LOGITS

 epilepsy

1.24

 epile

0.91

 patients

0.86

 dementia

0.81

ipolar

0.78

 mell

0.75

 cancer

0.75

 asthma

0.74

 disorder

0.74

 disorders

0.73

Activations Density 0.010%

medical conditions, particularly epilepsy and asthma

references to neurological and respiratory medical conditions

No Comments

No Known Activations