INDEX

Explanations

mentions of suicide

oai_token-act-pair · gpt-3.5-turbo

references to suicides and related topics

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

 plates

-0.73

 Plate

-0.73

 Lann

-0.72

lat

-0.71

Ùİ

-0.67

abba

-0.67

 cone

-0.67

box

-0.66

 Tune

-0.66

 Clean

-0.65

POSITIVE LOGITS

 suicides

3.57

 Suicide

2.80

 suicide

2.60

 suicidal

2.11

Su

1.59

icides

1.38

 antidepressants

1.32

icide

1.30

 antidepressant

1.29

 heroin

1.24

Activations Density 0.030%

mentions of suicide

references to suicides and related topics

No Comments

No Known Activations