INDEX

Explanations

Words related to work or working

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "work."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

anamo

-0.77

wcs

-0.70

ylon

-0.69

 Flavoring

-0.68

 constitu

-0.68

EStream

-0.68

 Cricket

-0.67

 Bour

-0.64

arta

-0.62

ĪĴ

-0.62

POSITIVE LOGITS

bench

1.16

 ethic

1.15

manship

1.07

station

1.07

work

1.04

flows

1.03

hops

0.95

works

0.83

fare

0.82

hardt

0.81

Activations Density 0.060%

Words related to work or working

occurrences of the word "work."

No Comments

No Known Activations