Search for: All records

Offline meta-reinforcement learning (meta-RL) aims to train agents that can generalize to unseen tasks using pre-collected data from related tasks. Recent approaches leverage the scalability of transformer architectures to model behavior sequences and support task adaptation using target task demonstrations. However, such data is often unavailable in real-world settings, where the task objective may be known but cannot be easily demonstrated. In contrast, humans routinely interpret and perform new tasks based solely on natural language instructions. In this work, we explore the potential of using natural language task descriptions to enable zero-shot task adaptation in offline meta-RL without requiring any data from the target task. We propose the Text-Guided Decision Transformer (TG-DT), a framework that enables zero-shot generalization by grounding policy learning in natural language. TG-DT learns a shared embedding space between task descriptions and behavioral trajectories via a dual contrastive and matching-based objective, ensuring robust alignment. A transformer-based policy is then conditioned on these aligned representations to generate task-appropriate actions. At test time, TG-DT synthesizes policies for unseen tasks using only their text descriptions and can optionally leverage a description-guided data sharing strategy to enhance adaptation. Experiments on standard offline meta-RL benchmarks, including MuJoCo and Meta-World, demonstrate that TG-DT achieves strong generalization to unseen tasks. 

Modern cities generate vast streams of urban dynamics data reflecting mobility demand, environmental conditions, and traffic patterns. The value of these data lies not only in individual modalities but in their integration—urban signals are highly interdependent, with changes in one modality often influencing others. Consequently, predicting any single urban dynamic requires information from multiple interrelated sources. Although numerous methods—ranging from deep learning models to recent LLM-based approaches—have been proposed, most are limited in scope. They either focus on single-modality prediction, rely on rigid model designs that lack flexibility, or overlook inter-modal dependencies. As a result, they struggle to adapt to dynamic urban conditions and suffer from degraded predictive performance across modalities. In this paper, we propose UniLLM, a unified large language model for multi-modal urban dynamics prediction. At its core, UniLLM introduces a Unified Cross-Modal Alignment Module that transforms heterogeneous urban data into latent representations while preserving modality-specific patterns and capturing cross-modal correlations through a contrastive learning objective. To support dynamic adaptation across tasks and modalities, we design a Routing-Aware Prompting Mechanism that learns soft prompts based on task context and modality semantics. Furthermore, a Multi-Modal Memory-Guided Adaptive Algorithm employs replay-based gradient coordination and Frank–Wolfe optimization to mitigate cross-modal catastrophic forgetting during fine-tuning. Extensive experiments across multiple cities and urban modalities demonstrate that UniLLM consistently outperforms state-of-the-art baselines. These results highlight UniLLM's potential as a flexible and robust forecasting model for real-world, multi-modal urban environments. 

Forest ecosystems play a critical role in the Earth system as major carbon sinks that are essential for carbon neutralization and climate change mitigation. However, the Earth has undergone significant deforestation and forest degradation, and the remaining forested areas are also facing increasing pressures from socioeconomic factors and climate change, potentially pushing them towards tipping points.Responding to the grand challenge, a theory-based Ecosystem Demography (ED) model has been continuously developed over the past two decades and serves as a key component in major initiatives, including the Global Carbon Budget, NASA Carbon Monitoring System, and US Greenhouse Gas Center. Despite its growing importance in combating climate change and shaping carbon policies, ED's expensive computation significantly limits its ability to estimate carbon dynamics at the global scale with high spatial resolution.Recently, machine learning (ML) models have shown promising potential in approximating theory-based models with interesting success in various domains including weather forecasting, thanks to the open-source benchmark datasets made available.However, there are currently no publicly available ML-ready datasets for global carbon dynamics forecasting in forest ecosystems. The limited data availability hinders the development of corresponding ML emulators. Furthermore, the inputs needed for running ED are highly complex with over a hundred variables from various remote sensing products. To bridge the gap, we develop a new ML-ready benchmark dataset, \textit{CarbonGlobe}, for carbon dynamics forecasting, featuring that: (1) the data has a global-scale coverage at 0.5 resolution; (2) the temporal range spans 40 years; (3) 

This paper introduces a novel approach to fine-tuning transformer-based models for various spatial-temporal downstream tasks. While fine-tuning approaches have shown remarkable success in fields like natural language processing, their efficacy in human-generated spatial-temporal data is often hindered by the complex spatial-temporal correlation and extensive reliance on labeled data. We introduce Knowledge Graph-Guided Spatial-Temporal Cross-task Fine-Tuning method, i.e., KG-STFT, a cross-task fine-tuning approach designed for adapting to data-scarce spatial-temporal tasks by leveraging rich knowledge embedded in similar downstream tasks and pre-trained model. Our KG-STFT framework utilizes i) a non-linear knowledge ensembler to capture and integrate knowledge embedded in different transformer blocks, and ii) constructs a task knowledge graph to "transfer" knowledge from data-rich to data-scarce tasks. Empirical experiments on real-world taxi trajectory data show that KG-STFT outperforms baselines, especially in data-scarce tasks, by leveraging task commonalities to improve fine-tuning. 

Given historical traffic distributions and associated urban conditions observed in a city, the conditional urban traffic estimation problem aims at estimating realistic future projections of the traffic under a set of new urban conditions, e.g., new bus routes, rainfall intensity, and travel demands. The problem is important in reducing traffic congestion, improving public transportation efficiency, and facilitating urban planning. However, solving this problem is challenging due to the strong spatial dependencies of traffic patterns and the complex relations between the traffic and urban conditions. Recently, we proposed a Complex-Condition-Controlled Generative Adversarial Network C3-GAN, which tackles both of the challenges and solves the urban traffic estimation problem under various complex conditions by adding a fixed embedding network and an inference network on top of the standard conditional GAN model. The randomly chosen embedding network transforms the complex conditions to latent vectors, and the inference network enhances the connections between the embedded vectors and the traffic data. However, a randomly chosen embedding network cannot always successfully extract features of complex urban conditions, which indicates C3-GAN is unable to uniquely map different urban conditions to proper latent distributions. Thus, C3-GAN would fail in certain traffic estimation tasks. Besides, C3-GAN is hard to train due to vanishing gradients and mode collapse problems. To address these issues, in this article, we extend our prior work by introducing a new deep generative model, namely, C3-GAN+, which significantly improves the estimation performance and model stability. C3-GAN+ has new objective, architecture, and training algorithm. The new objective applies Wasserstein loss to the conditional generation case to encourage stable training. Shared convolutional layers between the discriminator and the inference network help to capture spatial dependencies of traffic more efficiently, part of the shared convolutional layers are used to update the embedding network periodically aiming to encourage good representation and avoid model divergence. Extensive experiments on real-world datasets demonstrate that our C3-GAN+ produces high-quality traffic estimations and outperforms state-of-the-art baseline methods.