INDEX

Explanations

phrases related to instructions or steps

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

Maps

-0.71

Tag

-0.70

Gab

-0.67

 Grip

-0.67

AG

-0.66

Ä

-0.65

ADA

-0.65

è¡

-0.64

Ag

-0.63

Fab

-0.63

POSITIVE LOGITS

anwhile

0.81

 supra

0.73

 GOODMAN

0.72

unal

0.69

pite

0.68

bled

0.68

erity

0.68

upt

0.67

olean

0.67

 unrem

0.67

Activations Density 21.903%

phrases related to instructions or steps

No Comments

No Known Activations