INDEX

Explanations

names of individuals, specifically the name "Emily" in this case

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

ername

-0.56

xual

-0.53

lasses

-0.51

tenance

-0.49

stood

-0.49

kef

-0.49

emonium

-0.48

lay

-0.48

 specificity

-0.48

nesses

-0.47

POSITIVE LOGITS

 Dickinson

0.81

Lak

0.76

gdala

0.58

otte

0.57

pton

0.55

 gown

0.55

sburg

0.55

endi

0.53

ivered

0.50

issance

0.50

Activations Density 8.755%

names of individuals, specifically the name "Emily" in this case

No Comments

No Known Activations