Data Foundations of AI Community

This is the official X account for the Data Foundations of AI Community.

[Data Foundations of AI Community @DFAI_Community]

We are launching DFAI: 𝘿𝙖𝙩𝙖 𝙁𝙤𝙪𝙣𝙙𝙖𝙩𝙞𝙤𝙣𝙨 𝙤𝙛 𝘼𝙄:
🌐 https://t.co/p7tCcPJaUK

A Thread 🧵:

[Pingbang Hu 🇹🇼 @PingbangHu]

New Preprint Alert ⏰

Propose Dr. Post-training 🩺 a Data Regularization framework, making your data more effective with ZERO overheads

Experiments demonstrate faster training convergence across SFT, RLHF, RLVR over SOTA data selection, opening up new data optimization designs!

[Data Foundations of AI Community @DFAI_Community]

🚨 This Wednesday at DFAI seminar: Training Data Attribution as Explanations: Insights from User Studies by Elisa Nguyen!

🗓️ May 13, 2026
🕖 3 PM UTC | 11 AM ET
Don't miss it! 🔥

[Data Foundations of AI Community @DFAI_Community]

🚨 This Tuesday at DFAI seminar: Understanding Rollout Staleness & Selectivity for Efficient RL by Haizhong Zheng (@haizhong_zheng) on how to mitigate the rollout-side inefficiency!
🗓️ April 21, 2026
🕖 6 PM UTC | 2 PM ET
Don't miss it! 🔥

[Data Foundations of AI Community @DFAI_Community]

🗒️Reminder: This is happening tomorrow! Please save, share, and join us for this exciting talk!

[Data Foundations of AI Community @DFAI_Community]

🚨 This Friday at DFAI seminar: The Brittleness of AI Alignment — A Data & Rules Perspective
Luxi (Lucy) He (@LuxiHeLucy) on how benign data and well-intentioned rules can break alignment!
🗓️ April 10, 2026
🕖 7 PM UTC | 2 PM ET
Don't miss it! 🔥

[Data Foundations of AI Community @DFAI_Community]

🚨 This Friday at DFAI seminar: The Brittleness of AI Alignment — A Data & Rules Perspective
Luxi (Lucy) He (@LuxiHeLucy) on how benign data and well-intentioned rules can break alignment!
🗓️ April 10, 2026
🕖 7 PM UTC | 2 PM ET
Don't miss it! 🔥

[Data Foundations of AI Community @DFAI_Community]

🚀 Our second talk from the Data Foundations of AI (DFAI) seminar series is now live!
🎤 Watch it here: https://t.co/39EaT4GYcL
🔎 Learn more about the community: https://t.co/p7tCcPJaUK

[Data Foundations of AI Community @DFAI_Community]

🚨 Next Tuesday at DFAI seminar: Olmix — A Framework for Data Mixing Throughout LM Development
Mayee Chen (@MayeeChen) on data mixing for real-world LM training!
🗓️ Mar 24, 2026
🕖 6 PM UTC | 2 PM ET
Don't miss it! 🔥

[Data Foundations of AI Community @DFAI_Community]

🗒️Reminder: This is happening tomorrow! Please save, share, and join us for this exciting talk!

[Data Foundations of AI Community @DFAI_Community]

🚨 Next Tuesday at DFAI seminar: Olmix — A Framework for Data Mixing Throughout LM Development
Mayee Chen (@MayeeChen) on data mixing for real-world LM training!
🗓️ Mar 24, 2026
🕖 6 PM UTC | 2 PM ET
Don't miss it! 🔥

[Data Foundations of AI Community @DFAI_Community]

🚨 Next Tuesday at DFAI seminar: Olmix — A Framework for Data Mixing Throughout LM Development
Mayee Chen (@MayeeChen) on data mixing for real-world LM training!
🗓️ Mar 24, 2026
🕖 6 PM UTC | 2 PM ET
Don't miss it! 🔥

[Data Foundations of AI Community @DFAI_Community]

🚀 Our second talk from the Data Foundations of AI (DFAI) seminar series is now live!
🎤 Watch it here: https://t.co/v8OrKlKZfC
🔎 Learn more about the community: https://t.co/p7tCcPJaUK

[Data Foundations of AI Community @DFAI_Community]

🚨 Next Tuesday at DFAI seminar: Compute Optimal Scaling of Skills — Knowledge vs Reasoning

Nicholas Roberts (@nick11roberts) on skill-dependent scaling laws!

🗓️ Mar 10, 2026
🕖 7 PM UTC | 2 PM ET

Don't miss it! 🔥

[Data Foundations of AI Community @DFAI_Community]

🗒️Reminder: This is happening tomorrow! Please save, share, and join us for this exciting talk

‼️Note: Due to daylight saving, it's now 6 PM UTC!!

[Data Foundations of AI Community @DFAI_Community]

🚨 Next Tuesday at DFAI seminar: Compute Optimal Scaling of Skills — Knowledge vs Reasoning

Nicholas Roberts (@nick11roberts) on skill-dependent scaling laws!

🗓️ Mar 10, 2026
🕖 7 PM UTC | 2 PM ET

Don't miss it! 🔥

[Data Foundations of AI Community @DFAI_Community]

🚀 Our first talk from the Data Foundations of AI (DFAI) seminar series is now live!
🎤 Watch it here: https://t.co/k2nhKctUjX
🔎 Learn more about the community:
https://t.co/CQFjwzoJFM

[Data Foundations of AI Community @DFAI_Community]

🚨 Next Tuesday at DFAI seminar: More Fruitful SFT by Respecting The Learner’s Distribution

Dylan Zhang (@dylan_works_) on🍇 GRAPE & 🍐PEAR — policy-aware SFT that aligns supervision with the learner’s distribution.

🗓️ Mar 3, 2026
🕖 7 PM UTC | 2 PM ET

Don’t miss it! 🔥

[Data Foundations of AI Community @DFAI_Community]

@nick11roberts ℹ️For more information: https://t.co/lyxgapQXOB

[Data Foundations of AI Community @DFAI_Community]

🚨 Next Tuesday at DFAI seminar: Compute Optimal Scaling of Skills — Knowledge vs Reasoning

Nicholas Roberts (@nick11roberts) on skill-dependent scaling laws!

🗓️ Mar 10, 2026
🕖 7 PM UTC | 2 PM ET

Don't miss it! 🔥

[Data Foundations of AI Community @DFAI_Community]

🗒️Reminder: This is happening tomorrow!
Please save, share, and join us for this exciting talk 🔥

[Data Foundations of AI Community @DFAI_Community]

🚨 Next Tuesday at DFAI seminar: More Fruitful SFT by Respecting The Learner’s Distribution

Dylan Zhang (@dylan_works_) on🍇 GRAPE & 🍐PEAR — policy-aware SFT that aligns supervision with the learner’s distribution.

🗓️ Mar 3, 2026
🕖 7 PM UTC | 2 PM ET

Don’t miss it! 🔥

[Dylan Zhang @dylan_works_]

Thank you for having us — very happy to share what we’ve learned so far, and we’re eager to learn from the discussion!

[Data Foundations of AI Community @DFAI_Community]

@dylan_works_ ℹ️For more information:
https://t.co/lyxgapQXOB

[Data Foundations of AI Community @DFAI_Community]

🚨 Next Tuesday at DFAI seminar: More Fruitful SFT by Respecting The Learner’s Distribution

Dylan Zhang (@dylan_works_) on🍇 GRAPE & 🍐PEAR — policy-aware SFT that aligns supervision with the learner’s distribution.

🗓️ Mar 3, 2026
🕖 7 PM UTC | 2 PM ET

Don’t miss it! 🔥

[Data Foundations of AI Community @DFAI_Community]

Also tagging our advisory committee, thank you for your support! @danqi_chen @mirrokni @HazyResearch @dawnsongtweets @james_y_zou

[Data Foundations of AI Community @DFAI_Community]

Tagging our wonder co-organizers, please spread words! @wonderingishika @CaptainCuong @DengJunwei61721 @LuxiHeLucy @PingbangHu @huyuzheng1999 @zihengh1 @ruoxijia @li_tingwei1998 @ShixuanLiu40618 @Jiaqi_Ma_ @fredsala @lschmidt3 @evangelinejy99 @JiachenWang97

[Data Foundations of AI Community @DFAI_Community]

We are launching DFAI: 𝘿𝙖𝙩𝙖 𝙁𝙤𝙪𝙣𝙙𝙖𝙩𝙞𝙤𝙣𝙨 𝙤𝙛 𝘼𝙄:
🌐 https://t.co/p7tCcPJaUK

A Thread 🧵:

[Data Foundations of AI Community @DFAI_Community]

Finally, special thanks to all the Advisory Committee members!

[Data Foundations of AI Community @DFAI_Community]

We are launching DFAI: 𝘿𝙖𝙩𝙖 𝙁𝙤𝙪𝙣𝙙𝙖𝙩𝙞𝙤𝙣𝙨 𝙤𝙛 𝘼𝙄:
🌐 https://t.co/p7tCcPJaUK

A Thread 🧵:

[Data Foundations of AI Community @DFAI_Community]

This is a community efforts, and it's impossible without all the wonderful Organizers!