Searcharxiv⌕ Search

arXiv subjects

Leandros Tassiulas

Publications and source records attributed to Leandros Tassiulas.

At least 19 recordsLinked to original sources
arXiv2026

Contribution-Aware Bandwidth Allocation for Multimodal Split Learning

Multimodal models are increasingly the default option for perception at the network edge, yet they are trained almost entirely in the datacenter, because a client holding several sensor streams cannot host an encoder per modality. Split Learning makes such training feasible by keeping only the first layers on the device, at the cost of an uplink that must carry smashed activations for every modality at every step. Existing compression schemes give each modality the same keep-ratio, so the shared budget is divided in proportion to smashed-activation dimension, a quantity unrelated to how much each modality contributes to the fused prediction. We make that division an explicit decision and call it inter-modality allocation: under a fixed uplink budget, every policy transmits the same expected payload and differs only in how that payload is split across modalities. Our allocator, ModalShare, sets each modality's keep-ratio from a Shapley contribution score that the server computes over coalitions of activations it has already received. Measuring this score adds no uplink traffic and no client-side computation, and needs no prior knowledge of which stream is which. ModalShare improves accuracy over equal keep-ratios by 15.4 and 12.4 percentage points on CREMA-D and MVSA at matched payload in 5x compression, with strong performance across three compressors, three datasets, and four budgets. We show that existing compressors underperform in multimodal settings, with ModalShare recovering what gains are left behind.

cs.LG↗
arXiv2026

Restoring Incentive Compatibility in Two-Stage Energy Markets with Prosumers

A central challenge in modern energy market design is the formulation of a strategy-proof imbalance settlement layer that secures both the economic efficiency of the institution and the stability of the power grid. Public data reveals that the day-ahead market is strategically biased below actual consumer demand. Such empirical observations are explained by active prosumers which provide implementable incentives for demand under-reporting. Active prosumers buy energy in the day-ahead market and sell energy in the real-time market for balancing real-time energy deviations. By under-reporting their demand for the day ahead they inflate real-time imbalances and, under uniform pricing, they dispatch their generation assets more profitably. We model the two-stage institution under linear preferences and benchmark it against its associated competitive equilibria. We show that although consumers' incentives for demand under-reporting vanish when the day-ahead market scales, prosumers' incentives remain lower bounded by a positive gain which depends only on the real-time market generation stack and their shares over it. To restore incentive compatibility under the existing informational constraints, we design a leave-one-out contrastive scoring rule-based penalty that is implemented by the day-ahead market operator, incentivizes prosumers to report their demand truthfully and ensures small charges when participating honestly. We illustrate these results with numerical simulations on synthetic data and evaluate our mechanism on real-market data by first rationalizing demand reports as subjective equilibria of the induced game. Our mechanism demonstrates strong incentive alignment while retaining a low cost for honest participation.

cs.GT↗
arXiv2026

Reasoning through Verifiable Forecast Actions: Consistency-Grounded RL for Financial LLMs

Financial markets are characterized by extreme non-stationarity, low signal-to-noise ratios, and strong dependence on external information such as news, company fundamentals, and macroeconomic signals. Yet, existing approaches either abstract time-series into text or decouple forecasting from language-based reasoning, leading to a fundamental mismatch between qualitative reasoning and quantitative outcomes. To address this, we introduce StockR1, a time-series-enhanced LLM that unifies stock forecasting and financial reasoning through a verifiable forecast action. Based on a tool-call design, the model first emits a forecast action, which is a structured and interpretable representation of its qualitative market outlook. It then invokes a time-series decoder conditioned on this action to generate distributional future trajectories, leading to more informed question answering and financial reasoning. We optimize the full pipeline with reinforcement learning, where rewards jointly reflect answer validity, forecast accuracy, and consistency between generated actions and observed time-series dynamics. In addition, rewards are reweighted by a sample-level uncertainty scalar, encouraging the model to accommodate varying uncertainty in market dynamics. We evaluate StockR1 on financial question answering and stock forecasting over a large-scale 10-year benchmark. Our method consistently outperforms time-series baselines and general-purpose LLMs, improving reasoning accuracy by 17.7% (4B) and 25.9% (8B). These findings demonstrate that structuring the forecast actions establishes a powerful synergy between language reasoning and temporal prediction, enabling LLMs to reason through verifiable, interpretable, and numerically grounded decisions.

cs.LG↗
arXiv2026

Rank Is Not Capacity: Spectral Occupancy for Latent Graph Models

Graph representation learning has become a standard approach for analyzing networked data, with latent embeddings widely used for link prediction, community detection, and related tasks. Yet a basic design choice, the latent dimension, is still treated as a brittle hyperparameter, fixed before training and tuned by held-out performance. Learned factors are also identifiable only up to rotation and rescaling, so the nominal rank rarely coincides with the quantity that governs model behavior. We propose Spectral Prefix Extraction and Capacity-Targeted Representation Analysis (Spectra), which replaces rank as the unit of analysis with the spectrum of a learned positive semidefinite kernel, trace-normalized so that spectra are comparable across fits. The normalized eigenvalues form a distribution on the simplex, and their Shannon effective rank acts both as a summary of learned capacity and as a controllable training-time coordinate: a single scalar shapes this realized dimension during training, and bisection targets any desired value within the rank cap. To theoretically support that, we show local regularity and monotonicity of the realized-dimension profile. Across collaboration, social, biological, and infrastructure networks, Spectra traces performance--capacity frontiers that make the trade-off between predictive accuracy and realized dimension visible. It performs competitively with strong link-prediction baselines, yields aligned lower-capacity views of the same fitted model through spectral prefixes, and provides a principled handle on capacity in the overparameterized regime. Capacity thus becomes a property of the fitted model rather than a hyperparameter of the training.

cs.LG↗
arXiv2026

CLAD: A Clustered Label-Agnostic Federated Learning Framework for Joint Anomaly Detection and Attack Classification

The rapid expansion of the Internet of Things (IoT) and Industrial IoT (IIoT) has created a massive, heterogeneous attack surface that challenges traditional network security mechanisms. While Federated Learning (FL) offers a privacy-preserving alternative to centralized Intrusion Detection Systems (IDS), standard approaches struggle to generalize across diverse device behaviors and typically fail to utilize the vast amounts of unlabeled data present in realistic edge environments. To bridge these gaps, we propose CLAD, a holistic framework that seamlessly incorporates Clustered Federated Learning (CFL) with a novel Dual-Mode Micro-Architecture ($\text{DM}^2\text{A}$). This unified approach simultaneously tackles the two primary bottlenecks of IoT security: device heterogeneity and label scarcity. The $\text{DM}^2\text{A}$ component features a shared encoder followed by two branches, enabling joint unsupervised anomaly detection and supervised attack classification; this allows the framework to harvest intelligence from both labeled and unlabeled clients. Concurrently, the clustering component dynamically groups devices with congruent traffic patterns, preventing global model divergence. By carefully combining these elements, CLAD ensures that no data is discarded and distinct operational patterns are preserved. Extensive evaluations demonstrate that this integrated approach significantly outperforms state-of-the-art baselines, achieving a 30% relative improvement in detection performance in scenarios with 80% unlabeled clients, with only half the communication cost.

cs.LG↗
arXiv2026

Hierarchical Federated Learning for Networked AI: From Communication Saving to Architecture-Aware Design

Federated learning (FL) is fundamentally a distributed optimization problem executed by communicating agents with local data, local computation, and partial system visibility. Once FL is viewed through that lens, hierarchy is not merely a scalability mechanism. It becomes the natural place to rethink how distributed optimization should be organized over real multi-tier networks. This article argues that hierarchical federated learning (HFL) should move beyond its common framing as a communication-saving protocol and instead be viewed as an architecture-aware design framework for networked AI. The framework is organized around three coupled design axes: architectural parameters, layer-wise optimization decomposition, and layer-wise communication realization. The first axis determines the coordination geometry of learning through hierarchy depth, layer asymmetry, and layered connectivity. The second determines how the global FL objective is decomposed across layers and highlights modular multi-layer optimization as a major opportunity beyond one dominant method everywhere. The third determines how the distributed optimization is physically realized under heterogeneous communication regimes, from interference-limited lower tiers to reliable upper tiers. A central message is that, in HFL, convergence becomes architecture-dependent: it is directly shaped by the chosen hierarchy, the assigned optimization roles, and the communication mechanisms that connect them. We develop this viewpoint using large-scale wireless edge intelligence as a flagship networked AI setting, then provide a comparative perspective on flat FL, two-tier HFL, and deep HFL together with a regime-oriented design map. The resulting perspective positions HFL as a practical methodology for designing future networked AI systems.

cs.LG↗
arXiv2026

A Family of Open Time-Series Foundation Models for the Radio Access Network

The Radio Access Network (RAN) is evolving into a programmable and disaggregated infrastructure that increasingly relies on AI-native algorithms for optimization and closed-loop control. However, current RAN intelligence is still largely built from task-specific models tailored to individual functions, resulting in model fragmentation, limited knowledge sharing across tasks, poor generalization, and increased system complexity. To address these limitations, we introduce TimeRAN, a unified multi-task learning framework for time-series modeling in the RAN. TimeRAN leverages a lightweight time-series foundation model with few task-specific heads to learn transferable representations that can be efficiently adapted across diverse tasks with limited supervision. To enable large-scale pretraining, we further curate and open-source TimeRAN DataPile, the largest time-series corpus for RAN analytics to date, comprising over 355K time series and 0.56B measurements across diverse telemetry sources, protocol layers, and deployment scenarios. We evaluate TimeRAN across a comprehensive set of RAN analytics tasks, including anomaly detection, classification, forecasting, and imputation, and show that it achieves state-of-the-art performance with minimal or no task-specific fine-tuning. Finally, we integrate TimeRAN into a proof-of-concept 5G testbed and demonstrate that it operates efficiently with limited resource requirements in real-world scenarios.

cs.NI↗
arXiv2026

Benchmark for Assessing Olfactory Perception of Large Language Models

Here we introduce the Olfactory Perception (OP) benchmark, designed to assess the capability of large language models (LLMs) to reason about smell. The benchmark contains 1,010 questions across eight task categories spanning odor classification, odor primary descriptor identification, intensity and pleasantness judgments, multi-descriptor prediction, mixture similarity, olfactory receptor activation, and smell identification from real-world odor sources. Each question is presented in two prompt formats, compound names and isomeric SMILES, to evaluate the effect of molecular representations. Evaluating 21 model configurations across major model families, we find that compound-name prompts consistently outperform isomeric SMILES, with gains ranging from +2.4 to +18.9 percentage points (mean approx +7 points), suggesting current LLMs access olfactory knowledge primarily through lexical associations rather than structural molecular reasoning. The best-performing model reaches 64.4\% overall accuracy, which highlights both emerging capabilities and substantial remaining gaps in olfactory reasoning. We further evaluate a subset of the OP across 21 languages and find that aggregating predictions across languages improves olfactory prediction, with AUROC = 0.86 for the best performing language ensemble model. LLMs should be able to handle olfactory and not just visual or aural information.

cs.CL↗
arXiv2026

Rate-Fidelity Tradeoffs in All-Photonic and Memory-Equipped Quantum Switches

Quantum entanglement switches are a key building block for early quantum networks, and a central design question is whether near-term devices should use only flying photons or also incorporate quantum memories. We compare two architectures: an all-photonic entanglement generation switch (EGS) that repeatedly attempts Bell-state measurements (BSM) without storing qubits, and a quantum memory-equipped switch that buffers entanglement and triggers measurements only when heralded connectivity is available (herald-then-swap control). These two designs trade off simple, memoryless operation that avoids decoherence and memory-induced latency against heralding-based control that buffers entanglement to use BSMs more efficiently. We formalize both models under a common hardware abstraction and characterize their achievable rate-fidelity regions, yielding a benchmarking methodology that translates hardware and protocol parameters into network-level performance. Numerical evaluation quantifies the rate-fidelity tradeoffs of both models, identifies operating regions in which each architecture dominates, and shows how hardware and protocol knobs can be tuned to meet application-specific targets.

quant-ph↗
arXiv2026

LitBench: A Graph-Centric Large Language Model Benchmarking Tool For Literature Tasks

While large language models (LLMs) have become the de facto framework for literature-related tasks, they still struggle to function as domain-specific literature agents due to their inability to connect pieces of knowledge and reason across domain-specific contexts, terminologies, and nomenclatures. This challenge underscores the need for a tool that facilitates such domain-specific adaptation and enables rigorous benchmarking across literature tasks. To that end, we introduce LitBench, a benchmarking tool designed to enable the development and evaluation of domain-specific LLMs tailored to literature-related tasks. At its core, LitBench uses a data curation process that generates domain-specific literature sub-graphs and constructs training and evaluation datasets based on the textual attributes of the resulting nodes and edges. The tool is designed for flexibility, supporting the curation of literature graphs across any domain chosen by the user, whether high-level fields or specialized interdisciplinary areas. In addition to dataset curation, LitBench defines a comprehensive suite of literature tasks, ranging from node and edge level analyses to advanced applications such as related work generation. These tasks enable LLMs to internalize domain-specific knowledge and relationships embedded in the curated graph during training, while also supporting rigorous evaluation of model performance. Our results show that small domain-specific LLMs trained and evaluated on LitBench datasets achieve competitive performance compared to state-of-the-art models like GPT-4o and DeepSeek-R1. To enhance accessibility and ease of use, we open-source the tool along with an AI agent tool that streamlines data curation, model training, and evaluation.

cs.DL↗
arXiv2026

HypRAG: Hyperbolic Dense Retrieval for Retrieval Augmented Generation

Embedding geometry plays a fundamental role in retrieval quality, yet dense retrievers for retrieval-augmented generation (RAG) remain largely confined to Euclidean space. However, natural language exhibits hierarchical structure from broad topics to specific entities that Euclidean embeddings fail to preserve, causing semantically distant documents to appear spuriously similar and increasing hallucination risk. To address these limitations, we introduce hyperbolic dense retrieval, developing two model variants in the Lorentz model of hyperbolic space: HyTE-FH, a fully hyperbolic transformer, and HyTE-H, a hybrid architecture projecting pre-trained Euclidean embeddings into hyperbolic space. To prevent representational collapse during sequence aggregation, we introduce the Outward Einstein Midpoint, a geometry-aware pooling operator that provably preserves hierarchical structure. On MTEB, HyTE-FH outperforms equivalent Euclidean baselines, while on RAGBench, HyTE-H achieves up to 29% gains over Euclidean baselines in context relevance and answer relevance using substantially smaller models than current state-of-the-art retrievers. Our analysis also reveals that hyperbolic representations encode document specificity through norm-based separation, with over 20% radial increase from general to specific concepts, a property absent in Euclidean embeddings, underscoring the critical role of geometric inductive bias in faithful RAG systems.

cs.IR↗
arXiv2026

Fin-RATE: A Real-world Financial Analytics and Tracking Evaluation Benchmark for LLMs on SEC Filings

With the increasing deployment of Large Language Models (LLMs) in the finance domain, LLMs are increasingly expected to parse complex regulatory disclosures. However, existing benchmarks often focus on isolated details, failing to reflect the complexity of professional analysis that requires synthesizing information across multiple documents, reporting periods, and corporate entities. Furthermore, these benchmarks do not disentangle whether errors arise from retrieval failures, generation inaccuracies, domain-specific reasoning mistakes, or misinterpretation of the query or context, making it difficult to precisely diagnose performance bottlenecks. To bridge these gaps, we introduce Fin-RATE, a benchmark built on U.S. Securities and Exchange Commission (SEC) filings and mirroring financial analyst workflows through three pathways: detail-oriented reasoning within individual disclosures, cross-entity comparison under shared topics, and longitudinal tracking of the same firm across reporting periods. We benchmark 17 leading LLMs, spanning open-source, closed-source, and finance-specialized models, under both ground-truth context and retrieval-augmented settings. Results show substantial performance degradation, with accuracy dropping by 18.60% and 14.35% as tasks shift from single-document reasoning to longitudinal and cross-entity analysis. This degradation is associated with increased comparison hallucinations, temporal and entity mismatches, and is further reflected in declines in reasoning quality and factual consistency--limitations that existing benchmarks have yet to formally categorize or quantify.

cs.CE↗
arXiv2026

Multi-Modal Time Series Prediction via Mixture of Modulated Experts

Real-world time series exhibit complex and evolving dynamics, making accurate forecasting extremely challenging. Recent multi-modal forecasting methods leverage textual information such as news reports to improve prediction, but most rely on token-level fusion that mixes temporal patches with language tokens in a shared embedding space. However, such fusion can be ill-suited when high-quality time-text pairs are scarce and when time series exhibit substantial variation in characteristics, thus complicating cross-modal alignment. In parallel, mixture-of-experts (MoE) architectures have proven effective for both time series modeling and multi-modal learning, yet many existing MoE-based modality integration methods still depend on token-level fusion. To address this, we propose Expert Modulation, a new mechanism for multi-modal time series prediction that conditions both routing and expert computation on textual signals, enabling direct and efficient cross-modal control over expert behavior. Through theoretical analysis and experiments, our proposed method demonstrates strong improvements in multi-modal time series prediction. The current code implementation is available at https://github.com/BruceZhangReve/MoME

cs.LG↗
arXiv2025

Scheduling in Quantum Satellite Networks: Fairness and Performance Optimization

Quantum satellite networks offer a promising solution for achieving long-distance quantum communication by enabling entanglement distribution across global scales. This work formulates and solves the quantum satellite network scheduling problem by optimizing satellite-to-ground station pair assignments under realistic system and environmental constraints. Our framework accounts for limited satellite and ground station resources, fairness, entanglement fidelity thresholds, and real world non-idealities including atmospheric losses, weather and background noise. In addition, we incorporate the complexities of multi-satellite relays enabled via inter-satellite links. We propose an integer linear programming (ILP) based optimization framework that supports multiple scheduling objectives, allowing us to analyze tradeoffs between maximizing total entanglement distribution rate and ensuring fairness across ground station pairs. Our framework can also be used as a benchmark tool to measure the performance of other potential transmission scheduling policies.

quant-ph↗
arXiv2025

TelecomTS: A Multi-Modal Observability Dataset for Time Series and Language Analysis

Modern enterprises generate vast streams of time series metrics when monitoring complex systems, known as observability data. Unlike conventional time series from domains such as climate, observability data are zero-inflated, highly stochastic, and exhibit minimal temporal structure. Despite their importance, observability datasets remain underrepresented in public benchmarks due to proprietary restrictions and privacy concerns. Existing datasets are often anonymized and normalized, removing scale information and limiting their use for tasks such as anomaly detection, root cause analysis, and multi-modal reasoning. To address this gap, we introduce TelecomTS, a large-scale observability dataset derived from a 5G telecommunications network. TelecomTS features heterogeneous, de-anonymized covariates with explicit absolute scale information and provides a diverse suite of downstream tasks, including anomaly detection, root cause analysis, and multi-modal question-answering. Benchmarking state-of-the-art time series, language, reasoning, and multi-modal foundation models reveals that existing approaches struggle with the abrupt, noisy, and high-variance dynamics characteristic of observability data. Our experiments further underscore the importance of preserving covariates' absolute scale, emphasizing the need for foundation time series models that natively leverage scale information for practical real-world observability applications. The code is available at: https://github.com/Ali-maatouk/TelecomTS.

cs.AI↗
arXiv2025

SlicePilot: Demystifying Network Slice Placement in Heterogeneous Cloud Infrastructures

Cellular networks are comprised of software-based entities, with main functions encapsulated as Virtual Network Functions (VNFs) deployed on Commercial-off-the-Shelf (COTS) hardware. As a key enabler of 5G, network slicing offers logically isolated Quality of Service (QoS) for diverse use cases. With the transition to cloud-native infrastructures, optimizing network slice placement across multi-cloud environments remains challenging due to heterogeneous resource capabilities and varying slice-specific demands. This paper presents SlicePilot, a modular framework that enables autonomous and near-optimal VNF placement using a disaggregated Multi-Agent Reinforcement Learning (MARL) approach. SlicePilot collects real-world traffic profiles to estimate resource needs for each slice type. These estimates guide a MARL-based scheduler that minimizes deployment costs while satisfying QoS constraints. We evaluate SlicePilot on a multi-cloud testbed and demonstrate a 19x speed-up over combinatorial optimization methods, while keeping deployment costs within 7.8% of the optimal. Although SlicePilot results in 2.42x more QoS violations under high-load conditions, this trade-off is offset by faster decision-making and reduced computational overhead. Overall, SlicePilot delivers a scalable, cost-efficient solution for network slice placement, making it suitable for real-time deployments where responsiveness and efficiency are critical.

cs.NI↗
arXiv2025

5GC-Bench: A Framework for Stress-Testing and Benchmarking 5G Core VNFs

The disaggregated, cloud-native design of the 5G Core (5GC) enables flexibility and scalability but introduces significant challenges. Control-plane procedures involve complex interactions across multiple Virtual Network Functions (VNFs), while the user plane must sustain diverse and resource-intensive traffic. Existing tools often benchmark these dimensions in isolation, rely on synthetic workloads, or lack visibility into fine-grained resource usage. This paper presents 5GC-Bench, a modular framework for stress-testing the 5GC under realistic workloads. 5GC-Bench jointly emulates signaling and service traffic, supporting both VNF profiling and end-to-end service-chain analysis. By characterizing bottlenecks and resource demands, it provides actionable insights for capacity planning and performance optimization. We integrated 5GC-Bench with the OpenAirInterface (OAI) 5GC and deployed it on a real 5G testbed, demonstrating its ability to uncover resource constraints and expose cross-VNF dependencies under scenarios that mirror operational 5G deployments. To foster reproducibility and further research, we release publicly all the artifacts.

cs.NI↗
arXiv2025

D2Q Synchronizer: Distributed SDN Synchronization for Time Sensitive Applications

In distributed Software-Defined Networking (SDN), distributed SDN controllers require synchronization to maintain a global network state. Despite the availability of synchronization policies for distributed SDN architectures, most policies do not consider joint optimization of network and user performance. In this work, we propose a reinforcement learning-based algorithm called D2Q Synchronizer, to minimize long-term network costs by strategically offloading time-sensitive tasks to cost-effective edge servers while satisfying the latency requirements for all tasks. Evaluation results demonstrate the superiority of our synchronizer compared to heuristic and other learning policies in literature, by reducing network costs by at least 45% and 10%, respectively, while ensuring the QoS requirements for all user tasks across dynamic and multi-domain SDN networks.

cs.NI↗