B
Bobby Yan
Researcher at University of California, Berkeley
Publications - 4
Citations - 177
Bobby Yan is an academic researcher from University of California, Berkeley. The author has contributed to research in topics: Hindsight bias & Computer science. The author has an hindex of 1, co-authored 2 publications receiving 2 citations.
Papers
More filters
Journal ArticleDOI
Holistic Evaluation of Language Models
Percy Liang,Rishi Bommasani,Tony Lee,Dimitris Tsipras,Dilara Soylu,Michihiro Yasunaga,Yian Zhang,Deepak Narayanan,Yuhuai Wu,Ananya Kumar,Benjamin Newman,Binhang Yuan,Bobby Yan,Ce Zhang,Christian Cosgrove,Christopher D. Manning,Christopher R'e,Diana Acosta-Navas,Drew A. Hudson,Eric Zelikman,Esin Durmus,Faisal Ladhak,Frieda Rong,Hongyu Ren,Huaxiu Yao,Jue Wang,Keshav Santhanam,Laurel Orr,Lucia Zheng,Byron Rogers,Mirac M. Suzgun,Nathan S. Kim,Neel Guha,Niladri S. Chatterji,Peter Henderson,Qian Huang,Ryan Chi,Michael Xie,Shibani Santurkar,Surya Ganguli,Tatsunori Hashimoto,Thomas Icard,Tianyi Zhang,Vishrav Chaudhary,William Wang,Xuechen Li,Yifan Mai,Yuhui Zhang,Yuta Koreeda +48 more
TL;DR: The Holistic Evaluation of Language Models (HELM) as mentioned in this paper ) is a popular benchmark for language models, with 30 models evaluated on 16 core scenarios and 7 metrics, exposing important trade-offs.
Journal ArticleDOI
Hindsight Logging for Model Training
Rolando Garcia,Eric Liu,Vikram Sreekanti,Bobby Yan,Anusha Dandamudi,Joseph E. Gonzalez,Joseph M. Hellerstein,Koushik Sen +7 more
TL;DR: Flor, a record-replay system for hindsight logging in Python, is implemented and it is found that Flor replay achieves near-ideal scale-out and order-of-magnitude speedups in replay, with just 1.47% average runtime overhead from record.
Journal ArticleDOI
Hindsight logging for model training
Rolando Garcia,Eric Liu,Vikram Sreekanti,Bobby Yan,Anusha Dandamudi,Joseph E. Gonzalez,Joseph M. Hellerstein,Koushik Sen +7 more
TL;DR: In modern Machine Learning, model training is an iterative, experimental process that can consume enormous computation resources and developer time as discussed by the authors. To aid in that process, experienced model develo...
Journal Article
FORML: Learning to Reweight Data for Fairness
TL;DR: FORML improves equality of opportunity fairness criteria on image classification tasks, reduces bias of corrupted labels, and facilitates building more fair datasets via data condensation without pre-processing data or post-processing model outputs.