Related papers: Incremental Consistency Guarantees for Replicated …
A relational database is inconsistent if it does not satisfy a given set of integrity constraints. Nevertheless, it is likely that most of the data in it is consistent with the constraints. In this paper we apply logic programming based on…
RAG systems are increasingly deployed in high-stakes domains where users expect outputs to be consistent across semantically equivalent queries. However, existing systems often exhibit significant inconsistencies due to variability in both…
Since its introduction, the transformer has shifted the development trajectory away from traditional models (e.g., RNN, MLP) in time series forecasting, which is attributed to its ability to capture global dependencies within temporal…
Effective caching is crucial for the performance of modern-day computing systems. A key optimization problem arising in caching -- which item to evict to make room for a new item -- cannot be optimally solved without knowing the future.…
Many machine learning algorithms rely on iterative updates of uncertainty representations, ranging from variational inference and expectation-maximization, to reinforcement learning, continual learning, and multi-agent learning. In the…
The convergence properties of the Iterative water-filling (IWF) based algorithms have been derived in the ideal situation where the transmitters in the network are able to obtain the exact value of the interference plus noise (IPN)…
Plasticity and stability are needed in class-incremental learning in order to learn from new data while preserving past knowledge. Due to catastrophic forgetting, finding a compromise between these two properties is particularly challenging…
Reinforcement learning (RL) systems typically optimize scalar reward functions that assume precise and reliable evaluation of outcomes. However, real-world objectives--especially those derived from human preferences--are often uncertain,…
Recently, Transformer-based encoder-decoder models have demonstrated strong performance in multilingual speech recognition. However, the decoder's autoregressive nature and large size introduce significant bottlenecks during inference.…
Given the ever-increasing complexity of adaptable software systems and their commonly hidden internal information (e.g., software runs in the public cloud), machine learning based performance modeling has gained momentum for evaluating,…
Requirements traceability in safety-critical software development remains largely dependent on external documentation maintained separately from the systems it describes. This separation introduces structural fragility: traces degrade…
Cascade systems, consisting of a lightweight model processing all samples and a heavier, high-accuracy model refining challenging samples, have become a widely-adopted distributed inference approach to achieving high accuracy and…
The notion of replicable algorithms was introduced in Impagliazzo et al. [STOC '22] to describe randomized algorithms that are stable under the resampling of their inputs. More precisely, a replicable algorithm gives the same output with…
The ability to perform repeated Byzantine agreement lies at the heart of important applications such as blockchain price oracles or replicated state machines. Any such protocol requires the following properties: (1) \textit{Byzantine…
Test-time scaling improves the inference performance of Large Language Models (LLMs) but also incurs substantial computational costs. Although recent studies have reduced token consumption through dynamic self-consistency, they remain…
Efficiently querying data on embedded sensor and IoT devices is challenging given the very limited memory and CPU resources. With the increasing volumes of collected data, it is critical to process, filter, and manipulate data on the edge…
We prove that no fully transactional system can provide fast read transactions (including read-only ones that are considered the most frequent in practice). Specifically, to achieve fast read transactions, the system has to give up support…
Numerous distributed applications, such as cloud computing and distributed ledgers, necessitate the system to invoke asynchronous consensus objects an unbounded number of times, where the completion of one consensus instance is followed by…
Fast appearance variations and the distractions of similar objects are two of the most challenging problems in visual object tracking. Unlike many existing trackers that focus on modeling only the target, in this work, we consider the…
Cloud providers sell their idle capacity on markets through an auction-like mechanism to increase their return on investment. The instances sold in this way are called spot instances. In spite that spot instances are usually 90% cheaper…