

Sundai Nomads - Remote Hack
βMicro-hacks for your future AI paper - in one day.
βπ¬ Sundai Research is your day to try all the craziest research ideas you were thinking of but never had time to do.
β Start the day with a discussion of the hottest and most recent AI topics/papers/tutorials.
π₯ Tinker with it in small groups.
π§ Learn from and with each other.
π Ship a blog post by the end of the day!
βSundai Research is built for researchers who aim to publish to top AI conferences (general or field specific). Thorough understanding of ML theory and the nuances of AI model training is required.
βWhat is up with RL for LLMs?
βπ€ RL on 1 example? https://arxiv.org/abs/2504.20571
βπ€― RL on 0 examples? (i.e. random/incorrect reward) https://rethink-rlvr.notion.site/Spurious-Rewards-Rethinking-Training-Signals-in-RLVR-1f4df34dac1880948858f95aeb88872f
βπ΅βπ« RL without a reward? https://arxiv.org/abs/2505.19590 https://arxiv.org/abs/2505.20282
βπ§ Blog post scrutinizing reported results in these and a few other papers: https://safe-lip-9a8.notion.site/Incorrect-Baseline-Evaluations-Call-into-Question-Recent-LLM-RL-Claims-2012f1fbf0ee8094ab8ded1953c15a37. They argue that performance gains are smaller than reported, however still present for some of the methods.
βπ€ We'll start by reading/discussing these papers and reviewing their codebases. Then we'll choose a few to reproduce/explore using puzzles studied in one of our previous hacks: https://research.sundai.club/projects/613dd9fb-75c8-49c7-9583-d7997ad1e675.
βWe will provide GPUs for the day.
βLetβs hack!
βOur vision at Sundai Research:
β"Pioneering AI research through hacking."
βDeeply embedded in Harvard and MIT ecosystems, Sundai Research connects top AI researchers, creating opportunities for collaborative breakthroughs & future publications.
βLearn more: sundai.foundation
βOur goals each Sundai: (1) explore cutting-edge AI research problems (2) design and conduct experiments for measurable insights (3) catalyze future AI publications and innovations
βOur mission (1) grow exceptional AI researchers (2) advance AI science (2) spread the researcher mindset
βOur values:
βOpen. Balance. Shortcuts.
βOpen mind, open source, open to new ideas and people from all walks of life. This is how progress happens.
βBalance between serving our community and making big strides for the future. Balance between hard work and frolic. Balance between spending time with your loved ones vs pushing the envelop of what's possible.
βShortcut is an ingenious move that a research hacker makes to understand complex problems. Prioritize rapid experimentation that will provide actionable insights over prolonged deliberation.
βTerms of Participation:
βBy joining this event you agree to take care & vigilance of the following:
βleave the room as you found it. All whiteboards, chairs and tables must be cleaned and re-arranged as you found them. All trash must be in the bin.
βdo not let any strangers enter into any building behind you where sundai is taking place. if you see anyone you do not recognize who is suspicious, please let us know immediately.
βuse only the dedicated spaces to a sundai hack. if unsure - ask explicitly.
βbe respectful and considerate of others who may be in the building and are studying or working.
βif unsure - always ask.
βYou agree to these rules by coming to the event. Please read full terms carefully here: https://tinyurl.com/sundaiterms