CASTILLO: Characterizing Response Length Distributions of Large Language Models

Efficiently managing compute resources for Large Language Model (LLM) inference remains challenging due to the inherently stochastic and variable lengths of autoregressive text generation. Accurately estimating response lengths in advance enables proactive resource allocation, yet existing approaches either bias text generation towards certain lengths or rely on assumptions that ignore model- and prompt-specific variability. We introduce CASTILLO, a dataset characterizing response length distributions across 13 widely-used open-source LLMs evaluated on seven distinct instruction-following corpora. For each $\langle$prompt, model$\rangle$ sample pair, we generate 10 independent completions using fixed decoding hyper-parameters, record the token length of each response, and publish summary statistics (mean, std-dev, percentiles), along with the shortest and longest completions, and the exact generation settings. Our analysis reveals significant inter- and intra-model variability in response lengths (even under identical generation settings), as well as model-specific behaviors and occurrences of partial text degeneration in only subsets of responses. CASTILLO enables the development of predictive models for proactive scheduling and provides a systematic framework for analyzing model-specific generation behaviors. We publicly release the dataset and code to foster research at the intersection of generative language modeling and systems.

Robust Generalization of Graph Neural Networks for Carrier Scheduling

Battery-free sensor tags are devices that leverage backscatter techniques to communicate with standard IoT devices, thereby augmenting a network's sensing capabilities in a scalable way. For communicating, a sensor tag relies on an unmodulated carrier provided by a neighboring IoT device, with a schedule coordinating this provisioning across the network. Carrier scheduling--computing schedules to interrogate all sensor tags while minimizing energy, spectrum utilization, and latency--is an NP-Hard optimization problem. Recent work introduces learning-based schedulers that achieve resource savings over a carefully-crafted heuristic, generalizing to networks of up to 60 nodes. However, we find that their advantage diminishes in networks with hundreds of nodes, and degrades further in larger setups. This paper introduces RobustGANTT, a GNN-based scheduler that improves generalization (without re-training) to networks up to 1000 nodes (100x training topology sizes). RobustGANTT not only achieves better and more consistent generalization, but also computes schedules requiring up to 2x less resources than existing systems. Our scheduler exhibits average runtimes of hundreds of milliseconds, allowing it to react fast to changing network conditions. Our work not only improves resource utilization in large-scale backscatter networks, but also offers valuable insights in learning-based scheduling.

A Study of Continual Learning Under Language Shift

The recent increase in data and model scale for language model pre-training has led to huge training costs. In scenarios where new data become available over time, updating a model instead of fully retraining it would therefore provide significant gains. In this paper, we study the benefits and downsides of updating a language model when new data comes from new languages - the case of continual learning under language shift. Starting from a monolingual English language model, we incrementally add data from Norwegian and Icelandic to investigate how forward and backward transfer effects depend on the pre-training order and characteristics of languages, for different model sizes and learning rate schedulers. Our results show that, while forward transfer is largely positive and independent of language order, backward transfer can be either positive or negative depending on the order and characteristics of new languages. To explain these patterns we explore several language similarity metrics and find that syntactic similarity appears to have the best correlation with our results.

FMM-Head: Enhancing Autoencoder-based ECG anomaly detection with prior knowledge

Detecting anomalies in electrocardiogram data is crucial to identifying deviations from normal heartbeat patterns and providing timely intervention to at-risk patients. Various AutoEncoder models (AE) have been proposed to tackle the anomaly detection task with ML. However, these models do not consider the specific patterns of ECG leads and are unexplainable black boxes. In contrast, we replace the decoding part of the AE with a reconstruction head (namely, FMM-Head) based on prior knowledge of the ECG shape. Our model consistently achieves higher anomaly detection capabilities than state-of-the-art models, up to 0.31 increase in area under the ROC curve (AUROC), with as little as half the original model size and explainable extracted features. The processing time of our model is four orders of magnitude lower than solving an optimization problem to obtain the same parameters, thus making it suitable for real-time ECG parameters extraction and anomaly detection.

DeepGANTT: A Scalable Deep Learning Scheduler for Backscatter Networks

Recent backscatter communication techniques enable ultra low power wireless devices that operate without batteries while interoperating directly with unmodified commodity wireless devices. Commodity devices cooperate in providing the unmodulated carrier that the battery-free nodes need to communicate while collecting energy from their environment to perform sensing, computation, and communication tasks. The optimal provision of the unmodulated carrier limits the size of the network because it is an NP-hard combinatorial optimization problem. Consequently, previous works either ignore carrier optimization altogether or resort to suboptimal heuristics, wasting valuable energy and spectral resources. We present DeepGANTT, a deep learning scheduler for battery-free devices interoperating with wireless commodity ones. DeepGANTT leverages graph neural networks to overcome variable input and output size challenges inherent to this problem. We train our deep learning scheduler with optimal schedules of relatively small size obtained from a constraint optimization solver. DeepGANTT not only outperforms a carefully crafted heuristic solution but also performs within ~3% of the optimal scheduler on trained problem sizes. Finally, DeepGANTT generalizes to problems more than four times larger than the maximum used for training, therefore breaking the scalability limitations of the optimal scheduler and paving the way for more efficient backscatter networks.

Learning Combinatorial Optimization on Graphs: A Survey with Applications to Networking

Existing approaches to solving combinatorial optimization problems on graphs suffer from the need to engineer each problem algorithmically, with practical problems recurring in many instances. The practical side of theoretical computer science, such as computational complexity then needs to be addressed. Relevant developments in machine learning research on graphs is surveyed, for this purpose. We organize and compare the structures involved with learning to solve combinatorial optimization problems, with a special eye on the telecommunications domain and its continuous development of live and research networks.

Artificial Decision Making Under Uncertainty in Intelligent Buildings

Our hypothesis is that by equipping certain agents in a multi-agent system controlling an intelligent building with automated decision support, two important factors will be increased. The first is energy saving in the building. The second is customer value---how the people in the building experience the effects of the actions of the agents. We give evidence for the truth of this hypothesis through experimental findings related to tools for artificial decision making. A number of assumptions related to agent control, through monitoring and delegation of tasks to other kinds of agents, of rooms at a test site are relaxed. Each assumption controls at least one uncertainty that complicates considerably the procedures for selecting actions part of each such agent. We show that in realistic decision situations, room-controlling agents can make bounded rational decisions even under dynamic real-time constraints. This result can be, and has been, generalized to other domains with even harsher time constraints.

Modeling Epidemic Spread in Synthetic Populations - Virtual Plagues in Massively Multiplayer Online Games

A virtual plague is a process in which a behavior-affecting property spreads among characters in a Massively Multiplayer Online Game (MMOG). The MMOG individuals constitute a synthetic population, and the game can be seen as a form of interactive executable model for studying disease spread, albeit of a very special kind. To a game developer maintaining an MMOG, recognizing, monitoring, and ultimately controlling a virtual plague is important, regardless of how it was initiated. The prospect of using tools, methods and theory from the field of epidemiology to do this seems natural and appealing. We will address the feasibility of such a prospect, first by considering some basic measures used in epidemiology, then by pointing out the differences between real world epidemics and virtual plagues. We also suggest directions for MMOG developer control through epidemiological modeling. Our aim is understanding the properties of virtual plagues, rather than trying to eliminate them or mitigate their effects, as would be in the case of real infectious disease.

Anticipatory Guidance of Plot

An anticipatory system for guiding plot development in interactive narratives is described. The executable model is a finite automaton that provides the implemented system with a look-ahead. The identification of undesirable future states in the model is used to guide the player, in a transparent manner. In this way, too radical twists of the plot can be avoided. Since the player participates in the development of the plot, such guidance can have many forms, depending on the environment of the player, on the behavior of the other players, and on the means of player interaction. We present a design method for interactive narratives which produces designs suitable for the implementation of anticipatory mechanisms. Use of the method is illustrated by application to our interactive computer game Kaktus.