INDEX

Explanations

proper nouns, such as names of individuals, locations, and organizations

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

 weave

-0.57

 gauge

-0.55

icably

-0.55

 Gandhi

-0.55

icable

-0.55

 oppressed

-0.54

($)

-0.53

KKK

-0.53

 Ethiop

-0.53

POSITIVE LOGITS

idon

1.03

EStream

0.84

 Soup

0.76

stein

0.76

ornia

0.76

Fly

0.75

usky

0.74

cone

0.74

antine

0.72

glers

0.71

Activations Density 5.946%

proper nouns, such as names of individuals, locations, and organizations

No Comments

No Known Activations