RESEARCH

→ View original post on X — @montreal_ai

Measuring Data: Mitchell et al. Research on AI Evaluation

By

@montreal_ai

–

14 December 2022 15h23

Measuring Data Mitchell et al.: https://
arxiv.org/abs/2212.05129 #ArtificialIntelligence #DeepLearning #MachineLearning

→ View original post on X — @learnopencv

YOLOR Outperforms YOLOv7 on COCO Test-Dev Dataset

By

@learnopencv

–

14 December 2022 14h15

Did you know YOLOR performs better on the Coco test-dev set than YOLOv7?
YOLOR takes inspiration from how humans combine explicit & implicit knowledge to process previously unseen data.
▶️Check out our blog post to learn more.https://t.co/SUXpFNwMvj #yolor #computervision #ai pic.twitter.com/GcIyHlqr02
— Satya Mallick (@LearnOpenCV) 14 décembre 2022

Did you know YOLOR performs better on the Coco test-dev set than YOLOv7?
YOLOR takes inspiration from how humans combine explicit & implicit knowledge to process previously unseen data. Check out our blog post to learn more. https://
learnopencv.com/yolor-paper-ex
planation-inference-an-in-depth-analysis/
… #yolor #computervision #ai

Performance in ML means accuracy, not speed

By

–

14 December 2022 11h27

(For non-ML folks: “performance” means being right, not being fast — accuracy, F1 score, etc.)

ChatGPT sacrifices in-context learning for dialog context modeling

By

–

14 December 2022 11h15

“ChatGPT seems to have sacrificed nearly all of its in-context learning ability to trade for the ability to model dialog context.”

text-davinci-003 recovers in-context learning ability but worse than code-davinci-002

By

–

14 December 2022 11h15

“text-davinci-003 recovered (but still worse than code-davinci-002) some in-context learning ability that is lost in text-davinci-002 (presumably because it tunes the model with LM mix-in) and further improved zero-shot ability (thanks to RLHF).”

text-davinci-002 instruction tuning reduces in-context learning, boosts zero-shot ability

By

–

14 December 2022 11h15

“text-davinci-002, released in May-Jun 2022, is a supervised instruction-tuned model based on code-davinci-002. It is very likely that the instruction tuning on text-davinci-002 decreased the model’s in-context learning ability but increased the model’s zero-shot ability […]”

→ View original post on X — @sallyeaves

EU Green Capital Toolkit: Sustainable Cities Technology

By

@sallyeaves

–

14 December 2022 10h28

Making #cities more #sustainable New TOOLKIT live @EUGreenCapital https://
bit.ly/SDGsCities #Mobility #ESG #renewableenergy #SDGs @psb_dc #ClimateAction #SmartCities @helene_wpli @ipfconline1 @dinisguarda @kuriharan #research @SmartCityexpo @LSECities #data #AI #5G

History of GPT-3 and its emergent abilities explained

By

–

14 December 2022 10h06

You think ChatGPT is amazing — you’ve been hacking on computers for years, but this you can’t explain. How did we get here, and so suddenly? How does it know and do so much? @Francis_YAO_ of @EdinburghNLP explains the history of GPT-3: https://
yaofu.notion.site/How-does-GPT-O
btain-its-Ability-Tracing-Emergent-Abilities-of-Language-Models-to-their-Sources-b9a57ac0fcf74f30a1ab9e3e36fa1dc1
…