INDEX

Explanations

mentions of political leaders or figures by their official titles

oai_token-act-pair · gpt-3.5-turbo

references to prime ministers

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

anwhile

-0.84

pel

-0.72

ologies

-0.69

entric

-0.67

oldemort

-0.65

Ĥİ

-0.64

INESS

-0.64

okia

-0.63

 chalk

-0.63

paces

-0.63

POSITIVE LOGITS

 Minister

1.21

 ministerial

1.01

 minister

0.99

 Ministers

0.95

val

0.85

 ministers

0.84

knit

0.80

iary

0.78

 Election

0.76

Rib

0.75

Activations Density 0.017%

mentions of political leaders or figures by their official titles

references to prime ministers

No Comments

No Known Activations