INDEX

Explanations

references to specific names or titles mentioned in a longer text

oai_token-act-pair · gpt-3.5-turbo

New Auto-Interp

Configuration

neuronpedia/gpt2-small__res_scefr-ajt/6-res_scefr-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

ctica

-0.98

ITY

-0.95

lished

-0.90

ITAL

-0.87

hedral

-0.87

ity

-0.85

IAN

-0.83

today

-0.80

Ø§ÙĦ

-0.80

ertodd

-0.80

POSITIVE LOGITS

terday

1.14

asers

1.07

asing

1.06

asure

1.00

aser

1.00

velt

0.93

ldon

0.92

ases

0.91

oman

0.90

ats

0.90

Activations Density 0.810%

references to specific names or titles mentioned in a longer text

No Comments

No Known Activations