INDEX

Explanations

terms related to specific organizations and institutions, especially those involving research, health, and government

oai_token-act-pair · gpt-3.5-turbo

references to government or national institutions related to health and research

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

aldi

-0.85

DragonMagazine

-0.81

loe

-0.81

fleet

-0.74

forward

-0.70

neau

-0.69

bryce

-0.68

vind

-0.67

alore

-0.67

hor

-0.66

POSITIVE LOGITS

 Standards

0.85

 Health

0.76

 Corrections

0.75

 Obesity

0.73

 Obst

0.71

 Environmental

0.67

 Atmospheric

0.66

 Crime

0.65

 Pediatrics

0.64

 Certified

0.64

Activations Density 0.066%

terms related to specific organizations and institutions, especially those involving research, health, and government

references to government or national institutions related to health and research

No Comments

No Known Activations