Microsoft Research Lab – Asia
News & features
Abstracts: NeurIPS 2024 with Jindong Wang and Steven Euijong Whang
| Gretchen Huizinga, Jindong Wang, and Steven Euijong Whang
Researcher Jindong Wang and Associate Professor Steven Euijong Whang explore the NeurIPS 2024 work ERBench. ERBench leverages relational databases to create LLM benchmarks that can verify model rationale via keywords in addition to checking answer correctness.
We’re excited to be a part of #NeurIPS2024! Explore the future of AI with over 100 groundbreaking papers, including oral and spotlight sessions, on reinforcement learning, advanced language model training, and multilingual, culturally inclusive benchmarks.
Abstracts: NeurIPS 2024 with Weizhu Chen
| Amber Tingle and Weizhu Chen
Next-token prediction trains a language model on all tokens in a sequence. VP Weizhu Chen discusses his team’s 2024 NeurIPS paper on how distinguishing between useful and “noisy” tokens in pretraining can improve token efficiency and model performance.
Since the Industrial Revolution, the burning of fossil fuels and changes in land use, especially deforestation, have driven the rise in atmospheric carbon dioxide (CO2). While terrestrial vegetation and oceans serve as natural carbon sinks, absorbing some of this CO2,…
Although large language models (LLMs) excel in language-focused tasks like news writing, document summarization, customer service, and virtual assistants, they face challenges when it comes to learning and inference on numeric and structured industry data, such as tabular data and…
Research Focus: Week of December 2, 2024
Can a new SOS-RMT protocol enable more efficient CL-MPC?; A fair-by-design, cloud-based algorithmic trading platform; LLM2CLIP unlocks richer visual representation; New technique enhances Low-Rank Adaptation’s expressiveness, generalization capabilities.
MarS: A unified financial market simulation engine in the era of generative foundation models
| Weiqing Liu, Junjie Li, Yang Liu, Chang Xu, Shikai Fang, Lewen Wang, and Jiang Bian
Microsoft Research presents a new large market model and Financial Market Simulation Engine (MarS) to help empower financial researchers with enhanced efficiency and more accurate insights for downstream tasks in financial markets.
Awards | SIGMOBILE
Jie Xiong receives SIGMOBILE Test-of-Time Paper Award
ArrayTrack: A Fine-Grained Indoor Location System (USENIX NSDI, 2013) introduced a pioneering approach to RF-based indoor localization and wireless sensing by leveraging phase-array-based signal processing. Moving beyond traditional wireless signal strength measurements, it utilized multiple antennas and advocated for the…