David Nadeau

3.2K posts

David Nadeau banner
David Nadeau

David Nadeau

@Pythonner

Full cycle Machine Learning: I design, develop, deploy & support. Bot Populi, Bot Dei.

Ottawa, Canada Katılım Nisan 2009
570 Takip Edilen630 Takipçiler
François Chollet
François Chollet@fchollet·
Can you find the pattern? ■□□■ → ▣ ■□■■ → □ □■□■■ → □ ■□■■□■ → ▣ ■□■■■□■ → ▣ ■■■□■■■ → ▣ ■□□□■■■ → □ ■□■□□■■■ → □ □■■■□■■■□ → ▣ ■□■■□□■■□■ → ▣ ■□■■□□■■□■■ → ?
English
101
46
423
189.7K
Philippe J. Fournier
Philippe J. Fournier@338Canada·
“To what extent are you proud to be Canadian?” 🟢81% Very/Somewhat proud 🔴16% Not very/Not at all proud [Léger, June 23-25, 2023, n=1,512]
Philippe J. Fournier tweet media
English
61
66
311
126.5K
David Nadeau retweetledi
Raunak
Raunak@raunakdoesdev·
A recent work from @iddo claimed GPT4 can score 100% on MIT's EECS curriculum with the right prompting. My friends and I were excited to read the analysis behind such a feat, but after digging deeper, what we found left us surprised and disappointed. dub.sh/gptsucksatmit 🧵
Aran Komatsuzaki@arankomatsuzaki

Exploring the MIT Mathematics and EECS Curriculum Using Large Language Models Presents a comprehensive dataset of 4,550 questions and solutions from all MIT EECS courses required for obtaining a degree arxiv.org/abs/2306.08997

English
53
781
3.2K
3.5M
David Nadeau retweetledi
Mark Tenenholtz
Mark Tenenholtz@marktenenholtz·
Everyone should learn to fine-tune LLMs. You can try (and fail) to force GPT-4 on to every solution, or you can reduce your costs and improve your accuracy by fine-tuning a task-specific LLM. Here's why, and some tips for getting started:
English
22
90
813
270.3K
David Nadeau retweetledi
OpenAI
OpenAI@OpenAI·
We're launching ten $100,000 grants for building prototypes of a democratic process for steering AI. Our goal is to fund experimentation with methods for gathering nuanced feedback from everyone on how AI should behave. Apply by June 24, 2023: openai.com/blog/democrati…
English
521
1.1K
4K
1.9M
François Legault
François Legault@francoislegault·
En plus de mon balado, j’ai maintenant le plaisir de vous annoncer que je me lance sur Twitch! 🎮🐟
François Legault tweet media
Français
590
147
1.8K
567.1K
David Nadeau
David Nadeau@Pythonner·
@AlphaSignalAI If AI is the new electricity, training AI on AI-generated data is the mirage of perpetual motion...
English
0
0
1
142
Lior Alexander
Lior Alexander@LiorOnAI·
This is big news, ChatGPT just outperformed mechanical turk workers on text annotation tasks! We're getting closer to complete AI-based data annotation, which in turn, can be used to train AI models. It will cause a big shift in the industry. Paper: arxiv.org/pdf/2303.15056…
Lior Alexander tweet media
English
55
371
1.7K
393.6K
David Nadeau
David Nadeau@Pythonner·
@togelius Data scientists filing GPT-4 waitlist questionnaire
English
0
0
1
80
Julian Togelius
Julian Togelius@togelius·
Caption this image! The caption should refer to AI or the AI discourse
Julian Togelius tweet media
English
12
0
15
12.5K
David Nadeau
David Nadeau@Pythonner·
@rasbt I don't have a good understanding of RL so I was explaining it to myself as follow: annotation work for SL is slow & expensive = difficult to cover tasks and domains horizontally. RL is 'cheap' in comparison: thumbs up/down or answer ranking = allows covering general language.
English
1
0
9
7.6K
Sebastian Raschka
Sebastian Raschka@rasbt·
Why do latest language transformers (LLMs like ChatGPT etc.) use reinforcement learning (RL) for finetuning instead of regular supervised learning (SL)? There are at least 5 reasons ... [1/10]
English
32
304
2K
506.7K
Martin Laprise
Martin Laprise@mlaprise·
Here’s an idea: 1) Start a chatting session between ChatGPT and Google Bard about politics. 2) Broadcast it on Twitch. 3) Enjoy 🍿
English
1
0
1
138

Keşfet

@fchollet @338Canada @iddo @Guglielminetti @francoislegault @tasty_gigabyte7 @AlphaSignalAI @togelius