INDEX

Explanations

mentions of the word "Mort" with varying activations

oai_token-act-pair · gpt-3.5-turbo

references to the name "Mort" and its variations in different contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

manship

-0.76

DragonMagazine

-0.72

 Founders

-0.71

Ô

-0.69

FFFF

-0.69

âĻ¥

-0.68

lihood

-0.68

KEY

-0.67

 Ethiopia

-0.67

 Somali

-0.66

POSITIVE LOGITS

mort

1.28

imer

1.06

gage

0.99

ician

0.94

 mort

0.93

icia

0.92

heast

0.92

uary

0.92

ipop

0.89

arily

0.88

Activations Density 0.010%

mentions of the word "Mort" with varying activations

references to the name "Mort" and its variations in different contexts

No Comments

No Known Activations