INDEX

Explanations

questions being asked or conversations taking place involving individuals seeking information or explanations

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

unning

-1.03

\\\\

-0.96

abal

-0.85

\\\\\\\\

-0.83

ulic

-0.83

alde

-0.82

astered

-0.81

EStreamFrame

-0.81

audio

-0.79

 Brill

-0.79

POSITIVE LOGITS

 questions

1.00

 citizenship

0.96

 forgiveness

0.95

 hairc

0.95

Dad

0.94

 Majesty

0.93

perty

0.87

 pardon

0.87

 quizz

0.87

 farewell

0.87

Activations Density 0.442%

questions being asked or conversations taking place involving individuals seeking information or explanations

No Comments

No Known Activations