INDEX

Explanations

instances where someone is asked or requested to do something by another person

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

 Appears

-0.63

 Died

-0.61

requires

-0.60

etheless

-0.57

 Chatt

-0.56

alde

-0.56

HUD

-0.55

Yep

-0.55

quickShipAvailable

-0.55

productive

-0.55

POSITIVE LOGITS

 participate

1.09

 give

0.98

 reconsider

0.97

 perform

0.96

 accompany

0.95

 partake

0.95

 join

0.94

 nominate

0.94

 keep

0.93

 remove

0.93

Activations Density 10.597%

instances where someone is asked or requested to do something by another person

No Comments

No Known Activations