Related papers: Semantic Caching for OLAP via LLM-Based Query Cano…

Large language models (LLMs) with extended context windows enable powerful downstream applications but impose significant memory overhead, as caching all key-value (KV) states scales linearly with sequence length and batch size. Existing…

Text documents, including programs, typically have human-readable semantic structure. Historically, programmatic access to these semantics has required explicit in-document tagging. Especially in systems where the text has an execution…

This report investigates enhancing semantic caching effectiveness by employing specialized, fine-tuned embedding models. Semantic caching relies on embedding similarity rather than exact key matching, presenting unique challenges in…

Evaluating the reasoning abilities of large language models (LLMs) solely from final answers can obscure failures in intermediate steps, especially in multi-hop QA benchmarks without step-level annotations. To address this gap, we introduce…

Large language models (LLMs), endowed with exceptional reasoning capabilities, are adept at discerning profound user interests from historical behaviors, thereby presenting a promising avenue for the advancement of recommendation systems.…

Document images often have intricate layout structures, with numerous content regions (e.g. texts, figures, tables) densely arranged on each page. This makes the manual annotation of layout datasets expensive and inefficient. These…

Intent-driven Networks (IDNs) are crucial in enhancing network management efficiency by enabling the translation of high-level intents into executable configurations via a top-down approach. The escalating complexity of network…

The fundamental challenge of using Large Language Models (LLMs) for reliable, enterprise-grade analytics, such as sentiment prediction, is the conflict between the LLMs' inherent stochasticity (generative, non-deterministic nature) and the…

Accuracy-based evaluation cannot reliably distinguish genuine generalization from shortcuts like memorization, leakage, or brittle heuristics, especially in small-data regimes. In this position paper, we argue for mechanism-aware evaluation…

Retrieval-augmented generation and tool-integrated LLM agents increasingly depend on external textual sources. This reliance broadens the available attack surface, allowing adversaries to insert malicious instructions that trigger…

Novel contexts may often arise in complex querying scenarios such as in evidence-based medicine (EBM) involving biomedical literature, that may not explicitly refer to entities or canonical concept forms occurring in any fact- or rule-based…

Cooperative edge caching in overlapping zones couples Base Station (BS) decisions, making content replacement sensitive to spatial topology and temporal reuse. Conventional heuristics suffer from myopia, while Deep Reinforcement Learning…

We introduce benchmark signatures to characterize the capacity demands of LLM benchmarks and their overlaps. Signatures are sets of salient tokens from in-the-wild corpora whose model token perplexity, reflecting training exposure, predicts…

The amount of multidimensional data published on the semantic web (SW) is constantly increasing, due to initiatives such as Open Data and Open Government Data, among other ones. Models, languages, and tools, that allow to obtain valuable…

Tabular in-context learning (ICL) has recently achieved state-of-the-art (SOTA) performance on several tabular prediction tasks. Previously restricted to classification problems on small tables, recent advances such as TabPFN and TabICL…

Large language models (LLMs) have proven to be very capable, but access to frontier models currently relies on inference providers. This introduces trust challenges: how can we be sure that the provider is using the model configuration they…

Online Analytical Processing (OLAP) for relational databases is a business decision support application. The application receives queries about the business database, usually requesting to summarize many database records, and produces few…

Sampling is a basic operation in many inference-time algorithms of large language models (LLMs). To scale up inference efficiently with a limited compute, it is crucial to find an optimal allocation for sample compute budgets: Which…

Intelligent systems powered by large-scale sensor networks are shifting from predefined monitoring to intent-driven operation, revealing a critical Semantic-to-Physical Mapping Gap. While large language models (LLMs) excel at semantic…

3D semantic occupancy prediction offers an intuitive and efficient scene understanding and has attracted significant interest in autonomous driving perception. Existing approaches either rely on full supervision, which demands costly…