INDEX

Explanations

statements related to moral judgment and criticism towards others

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

inav

-0.59

osate

-0.56

ð

-0.54

urry

-0.53

ignment

-0.53

 Highlights

-0.53

 entails

-0.51

ESE

-0.50

 Highlands

-0.49

 Previous

-0.49

POSITIVE LOGITS

 willing

0.86

 able

0.85

 aware

0.83

 afraid

0.79

 unable

0.77

 incapable

0.77

 capable

0.76

 supposed

0.76

 immune

0.76

 happiest

0.76

Activations Density 16.114%

statements related to moral judgment and criticism towards others

No Comments

No Known Activations