INDEX

Explanations

phrases related to societal issues and personal experiences

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 overlook

-1.02

 roundup

-0.94

 inflic

-0.92

 incompet

-0.89

 drown

-0.88

 arch

-0.88

 qualified

-0.86

 shelling

-0.85

 adjud

-0.85

 aven

-0.84

POSITIVE LOGITS

They

1.46

We

1.46

Our

1.36

Everything

1.34

It

1.34

Sometimes

1.34

Too

1.28

There

1.28

When

1.27

What

1.27

Activations Density 0.535%

phrases related to societal issues and personal experiences

No Comments

No Known Activations