INDEX

Explanations

phrases related to doors and closed spaces

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

ting

-0.67

orp

-0.63

cies

-0.63

udeb

-0.62

TING

-0.62

uala

-0.61

ter

-0.61

TY

-0.60

Err

-0.59

icators

-0.59

POSITIVE LOGITS

 neighbor

0.83

steps

0.82

 neighbour

0.79

door

0.76

ħĭ

0.75

ways

0.69

wife

0.67

 pupil

0.64

 neighbors

0.64

pathy

0.62

Activations Density 7.491%

phrases related to doors and closed spaces

No Comments

No Known Activations