INDEX

Explanations

information relating to personal biographies, such as birthplaces, family relations, and careers

oai_token-act-pair · gpt-3.5-turbo

biographical details and professional background information

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

 takedown

-0.59

 deterrent

-0.58

 replay

-0.57

 deterrence

-0.56

existence

-0.55

>>

-0.55

Fact

-0.52

 whatsoever

-0.51

 harms

-0.51

metry

-0.50

POSITIVE LOGITS

 honorary

0.76

BuyableInstoreAndOnline

0.65

enegger

0.65

 stint

0.64

 organising

0.63

 assistants

0.63

 incorpor

0.63

 oversee

0.63

 befriend

0.62

 founding

0.62

Activations Density 2.585%

information relating to personal biographies, such as birthplaces, family relations, and careers

biographical details and professional background information

No Comments

No Known Activations