INDEX

Explanations

methods and methodologies used in various fields ranging from technology to creativity

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

igi

-0.62

 impression

-0.60

 Sons

-0.60

 juven

-0.58

 gloom

-0.56

ixture

-0.55

 challeng

-0.55

 Lumpur

-0.53

ments

-0.52

 Barkley

-0.52

POSITIVE LOGITS

ologies

1.45

ologically

1.09

ologic

1.05

ical

0.99

ology

0.99

OLOGY

0.99

ically

0.89

fare

0.85

sonian

0.84

ologist

0.82

Activations Density 9.874%

methods and methodologies used in various fields ranging from technology to creativity

No Comments

No Known Activations