INDEX

Explanations

phrases related to political news and government activities

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

Õ

-0.85

Ïī

-0.83

 Interstellar

-0.80

ocene

-0.79

 Horses

-0.79

 Wilde

-0.76

TPPStreamerBot

-0.74

ibaba

-0.74

adr

-0.73

uberty

-0.73

POSITIVE LOGITS

 aide

1.21

 adviser

1.06

 diplomat

1.05

 aides

1.03

 staffer

1.00

ranking

0.97

 advisor

0.97

 leadership

0.96

 negotiator

0.96

 prosecutor

0.95

Activations Density 8.364%

phrases related to political news and government activities

No Comments

No Known Activations