Related papers: Optimal Design of Lines Replaceable Units
Structured sparsity has emerged as a popular model pruning technique, widely adopted in various architectures, including CNNs, Transformer models, and especially large language models (LLMs) in recent years. A promising direction to further…
Large reasoning models (LRMs) achieve strong performance on complex reasoning tasks but often generate harmful responses to malicious user queries. This paper investigates the underlying cause of these safety risks and shows that the issue…
Distributed storage infrastructures require the use of data redundancy to achieve high data reliability. Unfortunately, the use of redundancy introduces storage and communication overheads, which can either reduce the overall storage…
Large language models (LLMs) deliver superior performance but require substantial computational resources and operate with relatively low efficiency, while smaller models can efficiently handle simpler tasks with fewer resources. LLM…
We introduce a novel kind of robustness in linear programming. A solution x* is called robust optimal if for all realizations of objective functions coefficients and constraint matrix entries from given interval domains there are…
To achieve systematic generalisation, it first makes sense to master simple tasks such as arithmetic. Of the four fundamental arithmetic operations (+,-,$\times$,$\div$), division is considered the most difficult for both humans and…
Complementarity problems, a class of mathematical optimization problems with orthogonality constraints, are widely used in many robotics tasks, such as locomotion and manipulation, due to their ability to model non-smooth phenomena (e.g.,…
Reproducibility of modeling is a problem that exists for any machine learning practitioner, whether in industry or academia. The consequences of an irreproducible model can include significant financial costs, lost time, and even loss of…
Minimizing leakage from computational states is a challenge when using many-level systems like superconducting quantum circuits as qubits. We realize and extend the quantum-hardware-efficient, all-microwave leakage reduction unit (LRU) for…
We study network design problems for nonlinear and nonconvex flow models without controllable elements under load scenario uncertainties, i.e., under uncertain injections and withdrawals. To this end, we apply the concept of adjustable…
A key strategy for balancing performance and cost in modern machine learning systems is to dynamically route queries to either a low-cost model or a more expensive oracle (such as a large pretrained model or human expert), an approach known…
In this paper, a novel linear algorithm is proposed for state estimation including bad data detection of power systems that are monitored both by conventional and synchrophasor measurements. Both types of data are treated simultaneously and…
Research efforts of the past fifty years have led to a development of linear integer programming as a mature discipline of mathematical optimization. Such a level of maturity has not been reached when one considers nonlinear systems subject…
Critical systems must be designed resilient to all kinds of malfunctions. We are especially interested by the loss of control authority over actuators. This malfunction considers actuators producing uncontrolled and possibly undesirable…
We study the Rectified Linear Unit (ReLU) dual, an existing dual formulation for stochastic programs that reformulates non-anticipativity constraints using ReLU functions to generate tight, non-convex, and mixed-integer representable cuts.…
Athletic robots demand a whole-body actuation system design that utilizes motors up to the boundaries of their performance. However, creating such robots poses challenges of integrating design principles and reasoning of practical design…
Large language model (LLM) inference performance is increasingly bottlenecked by the memory wall. While GPUs continue to scale raw compute throughput, they struggle to deliver scalable performance for memory bandwidth bound workloads. This…
Constructing first-principles models is usually a challenging and time-consuming task due to the complexity of the real-life processes. On the other hand, data-driven modeling, and in particular neural network models often suffer from…
Run-Length Encoding (RLE) is one of the most fundamental tools in data compression. However, its compression power drops significantly if there lacks consecutive elements in the sequence. In extreme cases, the output of the encoder may…
We consider the problem of designing experiments for the comparison of two regression curves describing the relation between a predictor and a response in two groups, where the data between and within the group may be dependent. In order to…