INDEX

Explanations

phrases related to examinations and investigations

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

gren

-0.61

 Hots

-0.55

activated

-0.54

activate

-0.54

gap

-0.54

âĹ¼

-0.53

externalActionCode

-0.51

blue

-0.49

fires

-0.49

finger

-0.48

POSITIVE LOGITS

ively

0.66

mble

0.65

 examinations

0.63

 examination

0.63

ugu

0.62

 Probe

0.62

 microsc

0.60

alus

0.60

ysis

0.59

ĸļ

0.58

Activations Density 12.547%

phrases related to examinations and investigations

No Comments

No Known Activations