GPT

Edited by Ruby, Multicore, Ben Pace, A_donor, et al. last updated 27th Aug 2022

GPT (Generative Pretrained Transformer) is a family of large transformer-based language models created by OpenAI. Its ability to generate remarkably human-like responses has relevance to discussions on AGI.

External links:

GPT-3 Paper

GPT-3 Website

Posts tagged GPT

89Collection of GPT-3 results

Ω

Kaj_Sotala

5y

Ω

24

70To what extent is GPT-3 capable of reasoning?

QΩ

TurnTrout, Daniel Kokotajlo

5y

QΩ

73

65GPT-3: a disappointing paper

nostalgebraist

5y

43

63GPT-3 Fiction Samples

gwern

5y

15

56$1000 bounty for OpenAI to show whether GPT3 was "deliberately" pretending to be stupider than it is

Ω

Bird Concept

5y

Ω

39

54Two Small Experiments on GPT-2

jimrandomh

7y

28

111Alignment As A Bottleneck To Usefulness Of GPT-3

Ω

johnswentworth

5y

Ω

57

72How "honest" is GPT-3?

QΩ

abramdemski, gwern

5y

QΩ

18

39'This Waifu Does Not Exist': 100,000 StyleGAN & GPT-2 samples

gwern

7y

6

37Does GPT-2 Understand Anything?

Douglas Summers-Stay

6y

23

37345M version GPT-2 released

[anonymous]6y

0

29Replicating the replication crisis with GPT-3?

skybrian

5y

10

19How well can the GPT architecture solve the parity task?

Q

FactorialCode, gwern

5y

Q

3

260larger language models may disappoint you [or, an eternally unfinished draft]

Ω

nostalgebraist

4y

Ω

31

140Developmental Stages of GPTs

Ω

orthonormal

5y

Ω

72

LESSWRONG
LW

GPT

LESSWRONG
LW

GPT