Ruoyu Sun
YOU?
Author Swipe
View article: Bridging Formal Language with Chain-of-Thought Reasoning to Geometry Problem Solving
Bridging Formal Language with Chain-of-Thought Reasoning to Geometry Problem Solving Open
Large vision language models exhibit notable limitations on Geometry Problem Solving (GPS) because of their unreliable diagram interpretation and pure natural-language reasoning. A recent line of work mitigates this by using symbolic solve…
View article: Second Language (Arabic) Acquisition of LLMs via Progressive Vocabulary Expansion
Second Language (Arabic) Acquisition of LLMs via Progressive Vocabulary Expansion Open
This paper addresses the critical need for democratizing large language models (LLM) in the Arab world, a region that has seen slower progress in developing models comparable to state-of-the-art offerings like GPT-4 or ChatGPT 3.5, due to …
View article: A Middle Path for On-Premises LLM Deployment: Preserving Privacy Without Sacrificing Model Confidentiality
A Middle Path for On-Premises LLM Deployment: Preserving Privacy Without Sacrificing Model Confidentiality Open
Privacy-sensitive users require deploying large language models (LLMs) within their own infrastructure (on-premises) to safeguard private data and enable customization. However, vulnerabilities in local environments can lead to unauthorize…
View article: PDHG-Unrolled Learning-to-Optimize Method for Large-Scale Linear Programming
PDHG-Unrolled Learning-to-Optimize Method for Large-Scale Linear Programming Open
Solving large-scale linear programming (LP) problems is an important task in various areas such as communication networks, power systems, finance and logistics. Recently, two distinct approaches have emerged to expedite LP solving: (i) Fir…
View article: When Expressivity Meets Trainability: Fewer than $n$ Neurons Can Work
When Expressivity Meets Trainability: Fewer than $n$ Neurons Can Work Open
Modern neural networks are often quite wide, causing large memory and computation costs. It is thus of great interest to train a narrower network. However, training narrow neural nets remains a challenging task. We ask two theoretical ques…
View article: Millimeter-Wave Channel-Sounder Performance Verification Using Vector Network Analyzer in a Controlled RF Channel
Millimeter-Wave Channel-Sounder Performance Verification Using Vector Network Analyzer in a Controlled RF Channel Open
A new comparison-to-reference performance verification technique compares an E-band channel-sounder and reference vector network analyzer measurements of the same controlled, static channel. This new technique reduces the number of inaccur…
View article: Landscape of Sparse Linear Network: A Brief Investigation.
Landscape of Sparse Linear Network: A Brief Investigation. Open
Network pruning, or sparse network has a long history and practical significance in modern application. A major concern for neural network training is that the non-convexity of the associated loss functions may cause bad landscape. We focu…
View article: Time Stability of Untethered Electronic Switched MIMO Millimeter-Wave Channel Sounders
Time Stability of Untethered Electronic Switched MIMO Millimeter-Wave Channel Sounders Open
Untethered channel sounders are used to measure mobile radio propagation channels. As high data rate, millimeter-wave and multiple-input multiple output (MIMO) are considered in the next generation communications systems, knowledge of the …
View article: Millimeter-Wave Radio Channels vs. Synthetic Beamwidth
Millimeter-Wave Radio Channels vs. Synthetic Beamwidth Open
High-gain narrow-beam antennas or beamformed antenna arrays will likely be used in millimeter-wave (mmWave) bands and 5G to mitigate the high path loss. Since many multipath components may be excluded by the narrow beam, the mmWave radio c…
View article: Measurement and Analysis of Channel Characteristics in Reflective Environments at 3.6 GHz and 14.6 GHz
Measurement and Analysis of Channel Characteristics in Reflective Environments at 3.6 GHz and 14.6 GHz Open
Recently, high frequency bands (above 6 GHz) have attracted more attention for the next generation communication systems due to the limited frequency resources below 6 GHz. To reveal the influence of frequency on propagation channels, chan…
View article: Impacts of Absorber Loadings on Simulating the Multipath Channel in a Reverberation Chamber
Impacts of Absorber Loadings on Simulating the Multipath Channel in a Reverberation Chamber Open
In reverberation chambers, the multipath channels with different delay characteristics can be generated by loading the chamber with different amounts of absorbers. This paper investigates the impacts of absorber loadings on the delay chara…