INDEX

Explanations

official titles and positions of authority in various organizations or entities

oai_token-act-pair · gpt-3.5-turbo

mentions of official titles and roles associated with law and government

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

 manoeuv

-0.66

 revolutions

-0.65

 countries

-0.64

 realised

-0.63

 onwards

-0.62

 realise

-0.62

 onward

-0.61

zbollah

-0.60

British

-0.60

 fulfil

-0.60

POSITIVE LOGITS

 Kathleen

1.27

 Shawn

1.24

 Stephanie

1.23

 Cheryl

1.23

 Melissa

1.21

 Larry

1.19

 Lori

1.18

 Darrell

1.18

 Deborah

1.17

 Jennifer

1.17

Activations Density 0.283%

official titles and positions of authority in various organizations or entities

mentions of official titles and roles associated with law and government

No Comments

No Known Activations