Asia
AI raises the stakes for national security. Here's how to get it right
AI regulation requires harmonization between state and federal approaches, but fragmented policies risk undermining US competitive advantage in this critical technology.
Mitigating Source Bias for Fairer Weak Supervision
Theoretically, we show that it is possible for our approach to simultaneously improve both accuracy and fairness--in contrast to standard fairness approaches that suffer from tradeoffs. Empirically, we show that our technique improves accuracy on weak supervision baselines by as much as 32% while reducing demographic parity gap by 82.5%.
A Careful Examination of Large Language Model Performance on Grade School Arithmetic
Further analysis suggests a positive relationship (Spearman's r
Distributed Low-rank Matrix Factorization With Exact Consensus
Zhihui Zhu, Qiuwei Li, Xinshuo Yang, Gongguo Tang, Michael B. Wakin
Neural Information Processing Systems http://nips.cc/
Zero-Shot Tokenizer Transfer Benjamin Minixhofer
Our method comes close to the original models' performance in cross-lingual and
Privacy-Preserving Classification of Personal Text Messages with Secure Multi-Party Computation
Devin Reich, Ariel Todoki, Rafael Dowsley, Martine De Cock, anderson nascimento
Neural Information Processing Systems http://nips.cc/
Semi-flat minima and saddle points by embedding neural networks to overparameterization
Kenji Fukumizu, Shoichiro Yamaguchi, Yoh-ichi Mototake, Mirai Tanaka
We theoretically study the landscape of the training error for neural networks in overparameterized cases.
2D-OOB: Attributing Data Contribution Through Joint Valuation Framework
Data valuation has emerged as a powerful framework for quantifying each datum's