Large Language Models (LLMs) have shown strong potential for tabular data generation by modeling textualized feature-value pairs. However, tabular data inherently exhibits sparse feature-level dependencies, where many feature interactions are structurally insignificant. This creates a fundamental mismatch as LLMs' self-attention mechanism inevitably distributes focus across all pairs, diluting attention on critical relationships, particularly in datasets with complex dependencies or semantically ambiguous features. To address this limitation, we propose GraDe (Graph-Guided Dependency Learning), a novel method that explicitly integrates sparse dependency graphs into LLMs' attention mechanism. GraDe employs a lightweight dynamic graph learning module guided by externally extracted functional dependencies, prioritizing key feature interactions while suppressing irrelevant ones. Our experiments across diverse real-world datasets demonstrate that GraDe outperforms existing LLM-based approaches by up to 12% on complex datasets while achieving competitive results with state-of-the-art approaches in synthetic data quality. Our method is minimally intrusive yet effective, offering a practical solution for structure-aware tabular data modeling with LLMs.

Due to climate-induced changes, many habitats are experiencing range shifts away from their traditional geographic locations (Piguet, 2011). We propose a solution to accurately model whether bird species are present in a specific habitat through the combination of Convolutional Neural Networks (CNNs) (O'Shea, 2015) and tabular data. Our approach makes use of satellite imagery and environmental features (e.g., temperature, precipitation, elevation) to predict bird presence across various climates. The CNN model captures spatial characteristics of landscapes such as forestation, water bodies, and urbanization, whereas the tabular method uses ecological and geographic data. Both systems predict the distribution of birds with an average accuracy of 85%, offering a scalable but reliable method to understand bird migration.

Large language models (LLMs) have demonstrated strong capabilities in translating natural language questions about relational databases into SQL queries. In particular, test-time scaling techniques such as Self-Consistency and Self-Correction can enhance SQL generation accuracy by increasing computational effort during inference. However, these methods have notable limitations: Self-Consistency may select suboptimal outputs despite majority votes, while Self-Correction typically addresses only syntactic errors. To leverage the strengths of both approaches, we propose CSC-SQL, a novel method that integrates Self-Consistency and Self-Correction. CSC-SQL selects the two most frequently occurring outputs from parallel sampling and feeds them into a merge revision model for correction. Additionally, we employ the Group Relative Policy Optimization (GRPO) algorithm to fine-tune both the SQL generation and revision models via reinforcement learning, significantly enhancing output quality. Experimental results confirm the effectiveness and generalizability of CSC-SQL. On the BIRD private test set, our 7B model achieves 71.72\% execution accuracy, while the 32B model achieves 73.67\%. The code has been open sourced at https://github.com/CycloneBoy/csc_sql.

Agents based on large language models leverage tools to modify environments, revolutionizing how AI interacts with the physical world. Unlike traditional NLP tasks that rely solely on historical dialogue for responses, these agents must consider more complex factors, such as inter-tool relationships, environmental feedback and previous decisions, when making choices. Current research typically evaluates agents via multi-turn dialogues. However, it overlooks the influence of these critical factors on agent behavior. To bridge this gap, we present an open-source and high-quality benchmark $C^3$-Bench. This benchmark integrates attack concepts and applies univariate analysis to pinpoint key elements affecting agent robustness. In concrete, we design three challenges: navigate complex tool relationships, handle critical hidden information and manage dynamic decision paths. Complementing these challenges, we introduce fine-grained metrics, innovative data collection algorithms and reproducible evaluation methods. Extensive experiments are conducted on 49 mainstream agents, encompassing general fast-thinking, slow-thinking and domain-specific models. We observe that agents have significant shortcomings in handling tool dependencies, long context information dependencies and frequent policy-type switching. In essence, $C^3$-Bench aims to expose model vulnerabilities through these challenges and drive research into the interpretability of agent performance. The benchmark is publicly available at https://github.com/TencentHunyuan/C3-Benchmark.

Multimodal time series forecasting is foundational in various fields, such as utilizing satellite imagery and numerical data for predicting typhoons in climate science. However, existing multimodal approaches primarily focus on utilizing text data to help time series forecasting, leaving the visual data in existing time series datasets untouched. Furthermore, it is challenging for models to effectively capture the physical information embedded in visual data, such as satellite imagery's temporal and geospatial context, which extends beyond images themselves. To address this gap, we propose physics-informed positional e ncoding ( PIPE), a lightweight method that embeds physical information into vision language models (VLMs). PIPE introduces two key innovations: (1) a physics-informed positional indexing scheme for mapping physics to positional IDs, and (2) a variant-frequency positional encoding mechanism for encoding frequency information of physical variables and sequential order of tokens within the embedding space. By preserving both the physical information and sequential order information, PIPE significantly improves multimodal alignment and forecasting accuracy. Through the experiments on the most representative and the largest open-sourced satellite image dataset, PIPE achieves state-of-the-art performance in both deep learning forecasting and climate domain methods, demonstrating superiority across benchmarks, including a 12% improvement in typhoon intensity forecasting over prior works. Our code is provided in the supplementary material.

Adverse weather conditions, particularly convective phenomena, pose significant challenges to Air Traffic Management, often requiring real-time rerouting decisions that impact efficiency and safety. This study introduces a 3-D Gaussian Mixture Model to predict long lead-time flight trajectory changes, incorporating comprehensive weather and traffic data. Utilizing high-resolution meteorological datasets, including convective weather maps and wind data, alongside traffic records, the model demonstrates robust performance in forecasting reroutes up to 60 minutes. The novel 3-D Gaussian Mixture Model framework employs a probabilistic approach to capture uncertainty while providing accurate forecasts of altitude, latitude, and longitude. Extensive evaluation revealed a Mean Absolute Percentage Error below 0.02 across varying lead times, highlighting the model's accuracy and scalability. By integrating explainability techniques such as the Vanilla Gradient algorithm, the study provides insights into feature contributions, showing that they contribute to improving Air Traffic Management strategies to mitigate weather-induced disruptions.

Flight trajectory prediction is a critical time series task in aviation. While deep learning methods have shown significant promise, the application of large language models (LLMs) to this domain remains underexplored. This study pioneers the use of LLMs for flight trajectory prediction by reframing it as a language modeling problem. Specifically, We extract features representing the aircraft's position and status from ADS-B flight data to construct a prompt-based dataset, where trajectory waypoints are converted into language tokens. The dataset is then employed to fine-tune LLMs, enabling them to learn complex spatiotemporal patterns for accurate predictions. Comprehensive experiments demonstrate that LLMs achieve notable performance improvements in both single-step and multi-step predictions compared to traditional methods, with LLaMA-3.1 model achieving the highest overall accuracy. However, the high inference latency of LLMs poses a challenge for real-time applications, underscoring the need for further research in this promising direction.

Mapping and navigation services like Google Maps, Apple Maps, Openstreet Maps, are essential for accessing various location-based data, yet they often struggle to handle natural language geospatial queries. Recent advancements in Large Language Models (LLMs) show promise in question answering (QA), but creating reliable geospatial QA datasets from map services remains challenging. We introduce MapQaTor, a web application that streamlines the creation of reproducible, traceable map-based QA datasets. With its plug-and-play architecture, MapQaTor enables seamless integration with any maps API, allowing users to gather and visualize data from diverse sources with minimal setup. By caching API responses, the platform ensures consistent ground truth, enhancing the reliability of the data even as real-world information evolves. MapQaTor centralizes data retrieval, annotation, and visualization within a single platform, offering a unique opportunity to evaluate the current state of LLM-based geospatial reasoning while advancing their capabilities for improved geospatial understanding. Evaluation metrics show that, MapQaTor speeds up the annotation process by at least 30 times compared to manual methods, underscoring its potential for developing geospatial resources, such as complex map reasoning datasets. The website is live at: https://mapqator.github.io/ and a demo video is available at: https://youtu.be/7_aV9Wmhs6Q.

Real-world enterprise text-to-SQL workflows often involve complex cloud or local data across various database systems, multiple SQL queries in various dialects, and diverse operations from data transformation to analytics. We introduce Spider 2.0, an evaluation framework comprising 632 real-world text-to-SQL workflow problems derived from enterprise-level database use cases. The databases in Spider 2.0 are sourced from real data applications, often containing over 1,000 columns and stored in local or cloud database systems such as BigQuery and Snowflake. We show that solving problems in Spider 2.0 frequently requires understanding and searching through database metadata, dialect documentation, and even project-level codebases. This challenge calls for models to interact with complex SQL workflow environments, process extremely long contexts, perform intricate reasoning, and generate multiple SQL queries with diverse operations, often exceeding 100 lines, which goes far beyond traditional text-to-SQL challenges. Our evaluations indicate that based on o1-preview, our code agent framework successfully solves only 17.0% of the tasks, compared with 91.2% on Spider 1.0 and 73.0% on BIRD. Our results on Spider 2.0 show that while language models have demonstrated remarkable performance in code generation -- especially in prior text-to-SQL benchmarks -- they require significant improvement in order to achieve adequate performance for real-world enterprise usage. Progress on Spider 2.0 represents crucial steps towards developing intelligent, autonomous, code agents for real-world enterprise settings. Our code, baseline models, and data are available at https://spider2-sql.github.io.

In August 2023, Scott Aaronson and I reported the results of testing GPT4 with the Wolfram Alpha and Code Interpreter plug-ins over a collection of 105 original high-school level and college-level science and math problems (Davis and Aaronson, 2023). In September 2024, I tested the recently released model GPT-4o1-preview on the same collection. Overall I found that performance had significantly improved, but was still considerably short of perfect. In particular, problems that involve spatial reasoning are often stumbling blocks. On September 12, OpenAI (2024) released two preliminary versions, "ChatGPT-o1-preview" and "ChatGPT-o1-mini" of a forthcoming product "ChatGPT-o1".