Yunyi Shen/็”ณไบ‘้€ธ ๐Ÿบ (@shenraphael) 's Twitter Profile
Yunyi Shen/็”ณไบ‘้€ธ ๐Ÿบ

@shenraphael

I do random statistics stuff at @MIT EECS. Once UW-Madison and PKU. Take wildlife photos. Opinions are mine.

ID: 1288542460831043585

linkhttp://yunyishen.github.io calendar_today29-07-2020 18:30:44

2,2K Tweet

2,2K Takipรงi

236 Takip Edilen

Yunyi Shen/็”ณไบ‘้€ธ ๐Ÿบ (@shenraphael) 's Twitter Profile Photo

Unfortunately I will have to miss ICML half due to internship half due to US visa situation. But if you want to talk about classic stats in LLM, Schrodinger bridge/diffusion in scientific applications or any other interest stuff please reach out!

Jenny Huang (@jennyhuang99) 's Twitter Profile Photo

๐ŸšจPast work shows: dropping just 0.1% of the data can change the conclusions of important studies. We show: Many approximations can fail to catch this. ๐Ÿ“ขCheck out our new TMLR paper (w/ David Burt, Yunyi Shen/็”ณไบ‘้€ธ ๐Ÿบ , Tin Nguyen, and Tamara Broderick ) ๐Ÿ‘‡ openreview.net/forum?id=m6EQ6โ€ฆ

Jenny Huang (@jennyhuang99) 's Twitter Profile Photo

๐ŸšจDropping Just a Handful of Preferences Can Change Top Large Language Model Rankings. ๐Ÿ“ Oral @ ICML 2025 Workshop on Models of Human Feedback for AI Alignment ๐Ÿ”—New work w/ Yunyi Shen/็”ณไบ‘้€ธ ๐Ÿบ, Dennis Wei, and Tamara Broderick openreview.net/forum?id=b9r1sโ€ฆ

๐ŸšจDropping Just a Handful of Preferences Can Change Top Large Language Model Rankings.
๐Ÿ“ Oral @ ICML 2025 Workshop on Models of Human Feedback for AI Alignment
๐Ÿ”—New work w/ <a href="/ShenRaphael/">Yunyi Shen/็”ณไบ‘้€ธ ๐Ÿบ</a>, Dennis Wei, and Tamara Broderick
openreview.net/forum?id=b9r1sโ€ฆ
Grant Sanderson (@3blue1brown) 's Twitter Profile Photo

Welp, somewhat embarrassingly, yesterday this account was briefly compromised and someone posing as me promoted a crypto project supposedly "tokenizing" manim. Thank you to those of you who quickly flagged it as spam and alerted me to the problem. For future reference, it's a

Hao Sun - IRLxLLM (@holarissun) 's Twitter Profile Photo

๐Ÿš€ RL is powering breakthroughs in LLM alignment, reasoning, and agentic apps. Are you ready to dive into the RL x LLM frontier? Join us at ACL 2025 ACLโ€™25 tutorial: Inverse RL Meets LLM Alignment this Sunday at Vienna๐Ÿ‡ฆ๐Ÿ‡น(Jul 27th, 9am) ๐Ÿ“„ Preprint at huggingface.co/papers/2507.13โ€ฆ

syz (@stephenz_y) 's Twitter Profile Photo

We found this out just today while already in the middle of putting supplementary figures together. Extremely frustrating given the already narrow scope and timelines for responding to reviews.