INDEX

Explanations

phrases indicating desires or intentions, such as wanting to do something or wanting something to happen

oai_token-act-pair · gpt-3.5-turbo

New Auto-Interp

Configuration

neuronpedia/gpt2-small__res_scefr-ajt/6-res_scefr-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

issance

-1.08

 Enhancement

-1.00

herer

-0.98

esse

-0.97

rawled

-0.96

hooting

-0.94

ortment

-0.92

utical

-0.90

ynthesis

-0.89

soType

-0.88

POSITIVE LOGITS

 anymore

1.54

 anybody

1.41

 anything

1.28

any

1.27

 anyone

1.25

 surprises

1.22

 necessarily

1.12

nor

1.09

 ever

1.08

 whatsoever

1.07

Activations Density 1.632%

phrases indicating desires or intentions, such as wanting to do something or wanting something to happen

No Comments

No Known Activations