INDEX

Explanations

Italian-sounding surnames

oai_token-act-pair · gpt-3.5-turbo

proper names, particularly those ending in "ucci."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

less

-0.71

venge

-0.69

lessness

-0.68

LESS

-0.66

 Scandinavian

-0.66

 Desire

-0.65

working

-0.64

 Norwegian

-0.64

mary

-0.64

selves

-0.62

POSITIVE LOGITS

ucci

1.25

cci

0.89

ulla

0.88

ieri

0.88

ourt

0.87

osi

0.84

osa

0.84

abba

0.82

otte

0.81

hetto

0.80

Activations Density 0.011%

Italian-sounding surnames

proper names, particularly those ending in "ucci."

No Comments

No Known Activations