INDEX

Explanations

words related to work and effort

oai_token-act-pair · gpt-3.5-turbo

New Auto-Interp

Configuration

neuronpedia/gpt2-small__res_scl-ajt/6-res_scl-ajt

Prompts (Dashboard)

12,288 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

46,080

Data Type

torch.float32

Hook Point

blocks.6.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

apollo-research/Skylion007-openwebtext-tokenizer-gpt2

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Ukrain

-1.20

wcs

-0.90

 constitu

-0.88

iren

-0.84

anamo

-0.80

EStream

-0.79

 champagne

-0.77

 Flavoring

-0.77

ylon

-0.75

 Bubble

-0.72

POSITIVE LOGITS

 ethic

1.70

flows

1.57

station

1.55

bench

1.43

horse

1.42

aday

1.36

manship

1.36

forces

1.18

tops

1.15

hops

1.14

Activations Density 5.333%

words related to work and effort

No Comments

No Known Activations