Deep learning (e.g., Transformer) has been widely and successfully used in multivariate time series forecasting (MTSF). Unlike existing methods that focus on training models from a single modal of time series input, large language models (LLMs) based MTSF methods with cross-modal text and time series input have recently shown great superiority, especially with limited temporal data. However, current LLM-based MTSF methods usually focus on adapting and fine-tuning LLMs, while neglecting the distribution discrepancy between textual and temporal input tokens, thus leading to sub-optimal performance. To address this issue, we propose a novel Cross-Modal LLM Fine-Tuning (CALF) framework for MTSF by reducing the distribution discrepancy between textual and temporal data, which mainly consists of the temporal target branch with temporal input and the textual source branch with aligned textual input. To reduce the distribution discrepancy, we develop the cross-modal match module to first align cross-modal input distributions. Additionally, to minimize the modality distribution gap in both feature and output spaces, feature regularization loss is developed to align the intermediate features between the two branches for better weight updates, while output consistency loss is introduced to allow the output representations of both branches to correspond effectively. Thanks to the modality alignment, CALF establishes state-of-the-art performance for both long-term and short-term forecasting tasks with low computational complexity, and exhibiting favorable few-shot and zero-shot abilities similar to that in LLMs. Code is available at \url{https://github.com/Hank0626/LLaTA}.

Time series prediction is crucial for understanding and forecasting complex dynamics in various domains, ranging from finance and economics to climate and healthcare. Based on Transformer architecture, one approach involves encoding multiple variables from the same timestamp into a single temporal token to model global dependencies. In contrast, another approach embeds the time points of individual series into separate variate tokens. The former method faces challenges in learning variate-centric representations, while the latter risks missing essential temporal information critical for accurate forecasting. In our work, we introduce GridTST, a model that combines the benefits of two approaches using innovative multi-directional attentions based on a vanilla Transformer. We regard the input time series data as a grid, where the $x$-axis represents the time steps and the $y$-axis represents the variates. A vertical slicing of this grid combines the variates at each time step into a \textit{time token}, while a horizontal slicing embeds the individual series across all time steps into a \textit{variate token}. Correspondingly, a \textit{horizontal attention mechanism} focuses on time tokens to comprehend the correlations between data at various time steps, while a \textit{vertical}, variate-aware \textit{attention} is employed to grasp multivariate correlations. This combination enables efficient processing of information across both time and variate dimensions, thereby enhancing the model's analytical strength. % We also integrate the patch technique, segmenting time tokens into subseries-level patches, ensuring that local semantic information is retained in the embedding. The GridTST model consistently delivers state-of-the-art performance across various real-world datasets.

Time series forecasting, the process of predicting future values based on observed historical data, is a pivotal task in various fields such as economics, finance, medicine, and environmental science. This task becomes particularly complex when dealing with multivariate temporal data, where predicting future values involves understanding the intricate interdependencies among multiple variables. This is essential in scenarios like weather forecasting, where variables such as temperature and humidity are interlinked, or in financial markets, where the stock prices of interconnected companies are observed. Recent advancements in time series forecasting have transitioned from conventional statistical approaches [1, 2] to sophisticated machine learning techniques, notably deep learning [3, 4, 5]. However, the field still grapples with a scarcity of both precise and interpretable models for multivariate time series prediction.

We present a new challenge to examine whether large language models understand social norms. In contrast to existing datasets, our dataset requires a fundamental understanding of social norms to solve. Our dataset features the largest set of social norm skills, consisting of 402 skills and 12,383 questions covering a wide set of social norms ranging from opinions and arguments to culture and laws. We design our dataset according to the K-12 curriculum. This enables the direct comparison of the social understanding of large language models to humans, more specifically, elementary students. While prior work generates nearly random accuracy on our benchmark, recent large language models such as GPT3.5-Turbo and LLaMA2-Chat are able to improve the performance significantly, only slightly below human performance. We then propose a multi-agent framework based on large language models to improve the models' ability to understand social norms. This method further improves large language models to be on par with humans. Given the increasing adoption of large language models in real-world applications, our finding is particularly important and presents a unique direction for future improvements.

Today's artificial intelligence is often described as a "black box." AI developers don't write explicit rules for these systems; instead, they feed in vast quantities of data and the systems learn on their own to spot patterns. But the inner workings of the AI models remain opaque, and efforts to peer inside them to check exactly what is happening haven't progressed very far. Beneath the surface, neural networks--today's most powerful type of AI--consist of billions of artificial "neurons" represented as decimal-point numbers. Nobody truly understands what they mean, or how they work.

Large language models (LLMs) have achieved remarkable performance on a wide range of tasks. However, recent studies have shown that LLMs can memorize training data and simple repeated tokens can trick the model to leak the data. In this paper, we take a step further and show that certain special characters or their combinations with English letters are stronger memory triggers, leading to more severe data leakage. The intuition is that, since LLMs are trained with massive data that contains a substantial amount of special characters (e.g. structural symbols {, } of JSON files, and @, # in emails and online posts), the model may memorize the co-occurrence between these special characters and the raw texts. This motivates us to propose a simple but effective Special Characters Attack (SCA) to induce training data leakage. Our experiments verify the high effectiveness of SCA against state-of-the-art LLMs: they can leak diverse training data, such as code corpus, web pages, and personally identifiable information, and sometimes generate non-stop outputs as a byproduct. We further show that the composition of the training data corpus can be revealed by inspecting the leaked data -- one crucial piece of information for pre-training high-performance LLMs. Our work can help understand the sensitivity of LLMs to special characters and identify potential areas for improvement.

This paper introduces alternators, a novel family of non-Markovian dynamical models for sequences. An alternator features two neural networks: the observation trajectory network (OTN) and the feature trajectory network (FTN). The OTN and the FTN work in conjunction, alternating between outputting samples in the observation space and some feature space, respectively, over a cycle. The parameters of the OTN and the FTN are not time-dependent and are learned via a minimum cross-entropy criterion over the trajectories. Alternators are versatile. They can be used as dynamical latent-variable generative models or as sequence-to-sequence predictors. When alternators are used as generative models, the FTN produces interpretable low-dimensional latent variables that capture the dynamics governing the observations. When alternators are used as sequence-to-sequence predictors, the FTN learns to predict the observed features. In both cases, the OTN learns to produce sequences that match the data. Alternators can uncover the latent dynamics underlying complex sequential data, accurately forecast and impute missing data, and sample new trajectories. We showcase the capabilities of alternators in three applications. We first used alternators to model the Lorenz equations, often used to describe chaotic behavior. We then applied alternators to Neuroscience, to map brain activity to physical activity. Finally, we applied alternators to Climate Science, focusing on sea-surface temperature forecasting. In all our experiments, we found alternators are stable to train, fast to sample from, yield high-quality generated samples and latent variables, and outperform strong baselines such as neural ODEs and diffusion models in the domains we studied.

Multivariate time series (MTS) forecasting has been extensively applied across diverse domains, such as weather prediction and energy consumption. However, current studies still rely on the vanilla point-wise self-attention mechanism to capture cross-variable dependencies, which is inadequate in extracting the intricate cross-correlation implied between variables. To fill this gap, we propose Variable Correlation Transformer (VCformer), which utilizes Variable Correlation Attention (VCA) module to mine the correlations among variables. Specifically, based on the stochastic process theory, VCA calculates and integrates the cross-correlation scores corresponding to different lags between queries and keys, thereby enhancing its ability to uncover multivariate relationships. Additionally, inspired by Koopman dynamics theory, we also develop Koopman Temporal Detector (KTD) to better address the non-stationarity in time series. The two key components enable VCformer to extract both multivariate correlations and temporal dependencies. Our extensive experiments on eight real-world datasets demonstrate the effectiveness of VCformer, achieving top-tier performance compared to other state-of-the-art baseline models. Code is available at this repository: https://github.com/CSyyn/VCformer.

Given a Large Language Model (LLM) generation, how can we identify which training data led to this generation? In this paper, we proposed RapidIn, a scalable framework adapting to LLMs for estimating the influence of each training data. The proposed framework consists of two stages: caching and retrieval. First, we compress the gradient vectors by over 200,000x, allowing them to be cached on disk or in GPU/CPU memory. Then, given a generation, RapidIn efficiently traverses the cached gradients to estimate the influence within minutes, achieving over a 6,326x speedup. Moreover, RapidIn supports multi-GPU parallelization to substantially accelerate caching and retrieval. Our empirical result confirms the efficiency and effectiveness of RapidIn.

Traffic flow prediction, a critical aspect of intelligent transportation systems, has been increasingly popular in the field of artificial intelligence, driven by the availability of extensive traffic data. The current challenges of traffic flow prediction lie in integrating diverse factors while balancing the trade-off between computational complexity and the precision necessary for effective long-range and large-scale predictions. To address these challenges, we introduce a Spatial-Temporal Selective State Space (ST-Mamba) model, which is the first to leverage the power of spatial-temporal learning in traffic flow prediction without using graph modeling. The ST-Mamba model can effectively capture the long-range dependency for traffic flow data, thereby avoiding the issue of over-smoothing. The proposed ST-Mamba model incorporates an effective Spatial-Temporal Mixer (ST-Mixer) to seamlessly integrate spatial and temporal data processing into a unified framework and employs a Spatial-Temporal Selective State Space (ST-SSM) block to improve computational efficiency. The proposed ST-Mamba model, specifically designed for spatial-temporal data, simplifies processing procedure and enhances generalization capabilities, thereby significantly improving the accuracy of long-range traffic flow prediction. Compared to the previous state-of-the-art (SOTA) model, the proposed ST-Mamba model achieves a 61.11\% improvement in computational speed and increases prediction accuracy by 0.67\%. Extensive experiments with real-world traffic datasets demonstrate that the \textsf{ST-Mamba} model sets a new benchmark in traffic flow prediction, achieving SOTA performance in computational efficiency for both long- and short-range predictions and significantly improving the overall efficiency and effectiveness of traffic management.