In this study, we address the language-guided time-series event detection problem: given a natural-language event description and a time series, the system returns event occurrences as temporal spans. Language-guided event detection leverages domain knowledge expressed in natural language to enable zero-shot detection without domain-specific fine-tuning. We introduce GroundingMOMENT, a zero-shot cross-modal grounding framework that aligns a time-series foundation model with a language model. GroundingMOMENT concurrently localizes both complex, high-level event occurrences and their constituent domain-agnostic structural primitives (e.g., spikes, plateaus). We further introduce GroundTSSyn, a scalable pipeline for generating synthetic time-series grounding data, and release GroundTSBench, the first open-source benchmark for language-guided time-series grounding. Trained exclusively on GroundTSBench, GroundingMOMENT outperforms time-series multimodal large language models and achieves fast and accurate event detection. Case studies further demonstrate its zero-shot generalizability to real-world data from diverse domains.

Time Series Event Detection (TSED) aims to localize semantically meaningful events in time series data, with critical applications in high-stakes domains. Unlike statistical anomalies, events are often defined by natural-language descriptions with internal temporal-logic structures across multiple physical channels. However, in real-world settings, dense event annotations are expensive to obtain, making purely supervised learning difficult. We introduce Language-guided TSED, a setting where a model is given textual event descriptions and must ground them to intervals in multivariate signals with little or no labeled data. To address this problem, we propose Event Logic Tree (ELT), a knowledge representation framework that converts linguistic descriptions into structured temporal logic over signal primitives. Building on ELT, we present SELA, a neuro-symbolic VLM agent framework that iteratively grounds primitives from signal visualizations and composes them under ELT constraints, producing both event intervals and faithful tree-structured explanations. We further release a real-world benchmark across energy and climate domains with expert knowledge and annotations. Experiments show that SELA improves over supervised fine-tuning and existing zero/few-shot time series reasoning baselines.

Designing emotionally intelligent conversational systems to provide comfort and advice to people experiencing distress is a compelling area of research. Recently, with advancements in large language models (LLMs), end-to-end dialogue agents without explicit strategy prediction steps have become prevalent. However, implicit strategy planning lacks transparency, and recent studies show that LLMs’ inherent preference bias towards certain socio-emotional strategies hinders the delivery of high-quality emotional support. To address this challenge, we propose decoupling strategy prediction from language generation, and introduce a novel dialogue strategy prediction framework, EmoDynamiX, which models the discourse dynamics between user fine-grained emotions and system strategies using a heterogeneous graph for better performance and transparency. Experimental results on two ESC datasets show EmoDynamiX outperforms previous state-of-the-art methods with a significant margin (better proficiency and lower preference bias). Our approach also exhibits better transparency by allowing backtracing of decision making.

Concevoir des systèmes conversationnels dotés d’une intelligence émotionnelle pour apporter du réconfort et des conseils aux personnes en détresse constitue un domaine de recherche particulièrement prometteur. Récemment, grâce aux avancées des grands modèles de langue (LLMs), les agents conversationnels entraînés de bout en bout sans étapes explicites de prédiction de stratégie de dialogue sont devenus plus courants. Cependant, la planification implicite de stratégie manque de transparence, et des études récentes montrent que la préférence inhérente des LLMs pour certaines stratégies socioémotionnelles nuit à la qualité du soutien émotionnel fourni. Pour relever ce défi, nous proposons de dissocier la prédiction de stratégies de la génération du langage et introduisons un nouveau cadre de prédiction de stratégie conversationnelle, EmoDynamiX, qui modélise la dynamique du discours entre les émotions fines du côté de l’utilisateur et les stratégies du système au moyen d’un graphe hétérogène, afin d’améliorer à la fois les performances et la transparence. Les résultats expérimentaux sur deux jeux de données de conversations pour le support émotionnel (ESC) montrent qu’EmoDynamiX surpasse de manière significative les méthodes antérieures à l’état de l’art (avec une meilleure maîtrise et un biais de préférence plus faible). Notre approche offre également une meilleure transparence en permettant de retracer le processus de prise de décision.