INDEX

Explanations

instances where someone is being notified or informed about something

oai_token-act-pair · gpt-3.5-turbo

frequent mentions of notifications and alerts related to events or incidents

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

 aspiration

-0.70

 highs

-0.68

 Vegas

-0.67

 clichÃ©

-0.66

 weakness

-0.65

Bol

-0.63

 emphasis

-0.63

 charisma

-0.62

 pedest

-0.62

 ambition

-0.62

POSITIVE LOGITS

 notified

3.58

 alerted

2.53

 notify

1.91

 contacted

1.84

 consulted

1.61

 briefed

1.60

 informed

1.56

 dispatched

1.34

 inspected

1.29

 instructed

1.25

Activations Density 0.012%

instances where someone is being notified or informed about something

frequent mentions of notifications and alerts related to events or incidents

No Comments

No Known Activations