INDEX

Explanations

medical, technical, and problem-solving terms and phrases

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

DragonMagazine

-0.75

SPONSORED

-0.73

Ã©e

-0.73

 Printed

-0.71

prints

-0.68

plates

-0.68

AppData

-0.68

thouse

-0.66

antry

-0.65

ateurs

-0.65

POSITIVE LOGITS

 deficiency

0.83

 problem

0.82

 deficiencies

0.81

 longstanding

0.81

 problems

0.80

 predicament

0.80

 equation

0.79

 satisf

0.78

missing

0.77

 dilemma

0.74

Activations Density 14.073%

medical, technical, and problem-solving terms and phrases

No Comments

No Known Activations