Related papers: Are Your Epochs Too Epic? Batch Free Can Be Harmfu…
We introduce a new class of exact Minimum-Bandwidth Regenerating (MBR) codes for distributed storage systems, characterized by a low-complexity uncoded repair process that can tolerate multiple node failures. These codes consist of the…
Model extraction aims to create a functionally similar copy from a machine learning as a service (MLaaS) API with minimal overhead, typically for illicit profit or as a precursor to further attacks, posing a significant threat to the MLaaS…
Online optimization with memory costs has many real-world applications, where sequential actions are made without knowing the future input. Nonetheless, the memory cost couples the actions over time, adding substantial challenges.…
Error Feedback (EF) is a highly popular and immensely effective mechanism for fixing convergence issues which arise in distributed training methods (such as distributed GD or SGD) when these are enhanced with greedy communication…
Elasticity is highly desirable for stream processing systems to guarantee low latency against workload dynamics, such as surges in data arrival rate and fluctuations in data distribution. Existing systems achieve elasticity following a…
A maximum distance separable (MDS) array code is composed of $m\times (k+r)$ arrays such that any $k$ out of $k+r$ columns suffice to retrieve all the information symbols. Expanded-Blaum-Roth (EBR) codes and Expanded-Independent-Parity…
Ad-hoc search calls for the selection of appropriate answers from a massive-scale corpus. Nowadays, the embedding-based retrieval (EBR) becomes a promising solution, where deep learning based document representation and ANN search…
A retrieval data structure stores a static function f : S -> {0,1}^r . For all x in S, it returns the r-bit value f(x), while for other inputs it may return an arbitrary result. The structure cannot answer membership queries, so it does not…
Mixture-of-Experts (MoE) models are typically pre-trained with explicit load-balancing constraints to ensure statistically balanced expert routing. Despite this, we observe that even well-trained MoE models exhibit significantly imbalanced…
When a computational task tolerates a relaxation of its specification or when an algorithm tolerates the effects of noise in its execution, hardware, programming languages, and system software can trade deviations from correct behavior for…
Real-world robotic agents must act under partial observability and long horizons, where key cues may appear long before they affect decision making. However, most modern approaches rely solely on instantaneous information, without…
Long chain-of-thought reasoning and agentic tool-calling produce traces spanning tens of thousands of tokens, yet Transformer KV caches grow linearly with sequence length, creating a memory bottleneck on commodity hardware. State-space…
Lock-free data structures are an important tool for the development of concurrent programs as they provide scalability, low latency and avoid deadlocks, livelocks and priority inversion. However, they require some sort of additional support…
In this paper, we first show that increases in beam size, even for small-sized LLMs (1B-7B params), require extensive GPU usage, leading to up to 80% of recurring crashes due to memory overloads in LLM-based APR. Seemingly simple solutions…
Operating systems have historically had to manage only a single type of memory device. The imminent availability of heterogeneous memory devices based on emerging memory technologies confronts the classic single memory model and opens a new…
Relative entropy coding (REC) algorithms encode a random sample following a target distribution $Q$, using a coding distribution $P$ shared between the sender and receiver. Sadly, general REC algorithms suffer from prohibitive encoding…
The rapid evolution of mobile edge computing (MEC) has introduced significant challenges in optimizing resource allocation in highly dynamic wireless communication systems, in which task offloading decisions should be made in real-time.…
The recently released persistent memory (PM) offers high performance, persistence, and is cheaper than DRAM. This opens up new possibilities for indexes that operate and persist data directly on the memory bus. Recent learned indexes…
As deep learning models continue to increase in size, the memory requirements for training have surged. While high-level techniques like offloading, recomputation, and compression can alleviate memory pressure, they also introduce…
Humans excel at lifelong learning, as the brain has evolved to be robust to distribution shifts and noise in our ever-changing environment. Deep neural networks (DNNs), however, exhibit catastrophic forgetting and the learned…