INDEX

Explanations

mentions of prominent individuals or notable figures

oai_token-act-pair · gpt-3.5-turbo

references to notable individuals, particularly in a contextual or relational manner

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

actionDate

-0.78

»Ĵ

-0.75

atform

-0.74

Enlarge

-0.69

manent

-0.69

sequent

-0.68

incial

-0.65

obos

-0.63

etheless

-0.61

farious

-0.61

POSITIVE LOGITS

 hates

1.63

 deserves

1.54

 loves

1.48

ain

1.44

 sucks

1.43

 wants

1.39

 knows

1.37

 owes

1.30

 thinks

1.30

 shouldn

1.30

Activations Density 0.369%

mentions of prominent individuals or notable figures

references to notable individuals, particularly in a contextual or relational manner

No Comments

No Known Activations