INDEX

Explanations

references to types of items used for comfort or safety

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

ctions

-0.59

ymes

-0.58

istry

-0.54

Naz

-0.53

erence

-0.53

ocratic

-0.52

uality

-0.52

ional

-0.52

ĸ

-0.51

orie

-0.51

POSITIVE LOGITS

 blanket

1.01

 blankets

0.72

abies

0.67

 COVER

0.60

 covers

0.58

 insulation

0.57

 knit

0.56

spread

0.56

 covering

0.56

 umbrella

0.55

Activations Density 9.754%

references to types of items used for comfort or safety

No Comments

No Known Activations

references to types of items used for comfort or safety

No Comments

No Known Activations