INDEX

Explanations

phrases related to global impact or reach

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

veyard

-0.67

yp

-0.66

romancer

-0.65

oga

-0.65

gre

-0.63

gotten

-0.63

nos

-0.62

gerald

-0.62

onies

-0.61

ippi

-0.60

POSITIVE LOGITS

 worldwide

0.76

wide

0.71

 wide

0.66

 acclaim

0.65

 renown

0.63

 dissemination

0.60

 warming

0.57

Wide

0.57

 ubiqu

0.57

 Wide

0.56

Activations Density 9.289%

phrases related to global impact or reach

No Comments

No Known Activations