INDEX

Explanations

references to business operations and transactions

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ussen

-0.46

ologies

-0.45

ilty

-0.45

antic

-0.45

 Trin

-0.44

obal

-0.43

urrection

-0.43

ellar

-0.43

uther

-0.42

ancies

-0.41

POSITIVE LOGITS

 droid

0.69

 dealings

0.62

 owners

0.59

 attire

0.55

 mogul

0.54

 ventures

0.52

 establishments

0.51

INESS

0.48

keepers

0.48

auer

0.46

Activations Density 5.924%

references to business operations and transactions

No Comments

No Known Activations