6 papers
HEPA: A Self-Supervised Horizon-Conditioned Event Predictive Architecture for Time Series
Jonas Petersen, Gian-Alessandro Lombardi, Riccardo Maggioni +3
Critical events in multivariate time series, from turbine failures to cardiac arrhythmias, demand accurate prediction, yet labeled data is scarce because such events are rare and c…
FactoryNet: A Large-Scale Dataset toward Industrial Time-Series Foundation Models
Karim Othman, Jonas Petersen, Matei Ignuta-Ciuncanu +5
We introduce the first universal pretraining corpus for industrial time-series data: FactoryNet. 51M datapoints across 23k end-to-end task executions (13.3k real, 9.8k synthetic) o…
What Structural Inductive Bias Helps Transformers Reason Over Knowledge Graphs? A Study with Tabula RASA
Jonas Petersen, Camilla Mazzoleni, Gian-Alessandro Lombardi +2
What structural inductive bias helps transformers reason over knowledge graphs? Through controlled ablations of a minimal transformer modification with four independently removable…
FactoryBench: Evaluating Industrial Machine Understanding
Yanis Merzouki, Coral Izquierdo, Matei Ignuta-Ciuncanu +8
We introduce FactoryBench, a benchmark for evaluating time-series models and LLMs on machine understanding over industrial robotic telemetry. Q&A pairs are organized along four cau…
Do Large Language Models Understand Word Senses?
Domenico Meconi, Simone Stirpe, Federico Martelli +2
Understanding the meaning of words in context is a fundamental capability for Large Language Models (LLMs). Despite extensive evaluation efforts, the extent to which LLMs show evid…
Truth or Mirage? Towards End-to-End Factuality Evaluation with LLM-Oasis
Alessandro Scirè, Andrei Stefan Bejgu, Simone Tedeschi +3
After the introduction of Large Language Models (LLMs), there have been substantial improvements in the performance of Natural Language Generation (NLG) tasks, including Text Summa…