INDEX

Explanations

bigram phrases that often refer to specific companies or 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

addons

-0.58

ority

-0.54

rem

-0.53

lahoma

-0.52

erella

-0.50

arat

-0.50

jas

-0.50

xit

-0.50

itivity

-0.50

ĸļ

-0.49

POSITIVE LOGITS

IZ

0.82

LECT

0.78

ibrary

0.76

INE

0.74

oaded

0.69

LE

0.68

TY

0.68

FORE

0.66

ESS

0.65

INK

0.64

Activations Density 6.144%

bigram phrases that often refer to specific companies or organizations

No Comments

No Known Activations