Deep learning and pre-trained models have shown great success in time series forecasting. However, in the tourism industry, time series data often exhibit a leading time property, presenting a 2D structure. This introduces unique challenges for forecasting in this sector. In this study, we propose a novel modelling paradigm, TripCast, which treats trip time series as 2D data and learns representations through masking and reconstruction processes. Pre-trained on large-scale real-world data, TripCast notably outperforms other state-of-the-art baselines in in-domain forecasting scenarios and demonstrates strong scalability and transferability in out-domain forecasting scenarios.

Aligning Large Language Models (LLM) to address subjectivity and nuanced preference levels requires adequate flexibility and control, which can be a resource-intensive and time-consuming procedure. Existing training-time alignment methods require full re-training when a change is needed and inference-time ones typically require access to the reward model at each inference step. To address these limitations, we introduce inference-time model alignment method that learns encoded representations of preference dimensions, called Alignment Vectors (AV). These representations are computed by subtraction of the base model from the aligned model as in model editing enabling dynamically adjusting the model behavior during inference through simple linear operations. Even though the preference dimensions can span various granularity levels, here we focus on three gradual response levels across three specialized domains: medical, legal, and financial, exemplifying its practical potential. This new alignment paradigm introduces adjustable preference knobs during inference, allowing users to tailor their LLM outputs while reducing the inference cost by half compared to the prompt engineering approach. Additionally, we find that AVs are transferable across different fine-tuning stages of the same model, demonstrating their flexibility. AVs also facilitate multidomain, diverse preference alignment, making the process 12x faster than the retraining approach. Aligning LLMs is crucial for adapting them to meet human preferences. Standard training-time alignment methods, such as RLHF (Ouyang et al., 2022) and DPO (Rafailov et al., 2024), are conducted during model training. However, making nuanced preference adjustments during inference with these approaches would necessitate retraining, which requires substantial amounts of time, preference data and computational resources. Inference-time LLM alignment, by contrast, delays the alignment process until inference (Wang et al., 2024). While preference alignment can be achieved through training-time methods or targeted prompting, fine-grained control over preferences at inference remains largely unexplored in current State-of-the-Art (SOTA) works (Sahoo et al., 2024; Guo et al., 2024).

Those methods usually rely on data curation reflecting on the deliberate reasoning path in specific application Large Language Models (LLMs) have recently demonstrated impressive areas. When it comes to complex application constraints, capabilities across various real-world applications. However, high-quality solutions often demand a large volume of data to cover due to the current text-in-text-out paradigm, it remains challenging enough data cases and the corresponding reasoning logic. This for LLMs to handle dynamic and complex application constraints, process fundamentally differs from the typical human cognition let alone devise general solutions that meet predefined process: faced with unfamiliar problems, people first seek to capture system goals. Current common practices like model finetuning and the overview of the underlying application constraints, which are reflection-based reasoning often address these issues case-by-case, potential rules summarized from observations. Next, these learned limiting their generalizability. To address this issue, we propose rules will be further refined when exception cases arise. The essence a flexible framework that enables LLMs to interact with system of this cognitive process lies in distilling rules and identifying minimal interfaces, summarize constraint concepts, and continually optimize cases for refinement rather than depending on the inefficient performance metrics by collaborating with human experts.

This paper identifies and analyzes applications in which Large Language Models (LLMs) can make Internet of Things (IoT) networks more intelligent and responsive through three case studies from critical topics: DDoS attack detection, macroprogramming over IoT systems, and sensor data processing. Our results reveal that the GPT model under few-shot learning achieves 87.6% detection accuracy, whereas the fine-tuned GPT increases the value to 94.9%. Given a macroprogramming framework, the GPT model is capable of writing scripts using high-level functions from the framework to handle possible incidents. Moreover, the GPT model shows efficacy in processing a vast amount of sensor data by offering fast and high-quality responses, which comprise expected results and summarized insights. Overall, the model demonstrates its potential to power a natural language interface. We hope that researchers will find these case studies inspiring to develop further.

Large language models (LLMs) are trained on vast amounts of data to generate natural language, enabling them to perform tasks like text summarization and question answering. These models have become popular in artificial intelligence (AI) assistants like ChatGPT and already play an influential role in how humans access information. However, the behavior of LLMs varies depending on their design, training, and use. In this paper, we uncover notable diversity in the ideological stance exhibited across different LLMs and languages in which they are accessed. We do this by prompting a diverse panel of popular LLMs to describe a large number of prominent and controversial personalities from recent world history, both in English and in Chinese. By identifying and analyzing moral assessments reflected in the generated descriptions, we find consistent normative differences between how the same LLM responds in Chinese compared to English. Similarly, we identify normative disagreements between Western and non-Western LLMs about prominent actors in geopolitical conflicts. Furthermore, popularly hypothesized disparities in political goals among Western models are reflected in significant normative differences related to inclusion, social inequality, and political scandals. Our results show that the ideological stance of an LLM often reflects the worldview of its creators. This raises important concerns around technological and regulatory efforts with the stated aim of making LLMs ideologically `unbiased', and it poses risks for political instrumentalization.

Leveraging large language models (LLMs) has garnered increasing attention and introduced novel perspectives in time series classification. However, existing approaches often overlook the crucial dynamic temporal information inherent in time series data and face challenges in aligning this data with textual semantics. To address these limitations, we propose HiTime, a hierarchical multi-modal model that seamlessly integrates temporal information into LLMs for multivariate time series classification (MTSC). Our model employs a hierarchical feature encoder to capture diverse aspects of time series data through both data-specific and task-specific embeddings. To facilitate semantic space alignment between time series and text, we introduce a dual-view contrastive alignment module that bridges the gap between modalities. Additionally, we adopt a hybrid prompting strategy to fine-tune the pre-trained LLM in a parameter-efficient manner. By effectively incorporating dynamic temporal features and ensuring semantic alignment, HiTime enables LLMs to process continuous time series data and achieves state-of-the-art classification performance through text generation. Extensive experiments on benchmark datasets demonstrate that HiTime significantly enhances time series classification accuracy compared to most competitive baseline methods. Our findings highlight the potential of integrating temporal features into LLMs, paving the way for advanced time series analysis. The code is publicly available for further research and validation. Our codes are publicly available1.

Long-context models (LCMs) have shown great potential in processing long input sequences (even more than 100M tokens) conveniently and effectively. With significant progress, recent research has pointed out that LCMs can accurately locate token-level salient information within the context. Yet, the generation performance of these LCMs is far from satisfactory and might result in misaligned responses, such as hallucinations. To enhance the generation capability of LCMs, existing works have investigated the effects of data size and quality for both pretraining and instruction tuning. Though achieving meaningful improvement, previous methods fall short in either effectiveness or efficiency. In this paper, we introduce LOGO (Long cOntext aliGnment via efficient preference Optimization), a training strategy that first introduces preference optimization for long-context alignment. To overcome the GPU memory-bound issue caused by the long sequence, LOGO employs a reference-free preference optimization strategy and adopts a position synthesis method to construct the training data. By training with only 0.3B data on a single 8 A800 GPU machine for 16 hours, LOGO allows the Llama-3-8B-Instruct-80K model to achieve comparable performance with GPT-4 in real-world long-context tasks while preserving the model's original capabilities on other tasks, e.g., language modeling and MMLU. Moreover, LOGO can extend the model's context window size while enhancing its generation performance. Figure 1: (a) Performance of LCMs on real-world long-context tasks; (b) Retrieval score (longcontext understanding ability) and recall score (generation ability) of LCMs on the synthetic retrieval long-context task (multi-value NIAH); (c) Long-context (pre-)training data size for each LCM.

The development of large language models (LLMs) has significantly expanded model sizes, resulting in substantial GPU memory requirements during inference. Nowadays, most existing KV cache compression methods focus on intra-layer compression within a single Transformer layer but few works consider layer-wise compression. In this paper, we propose a plug-and-play method called KVSharer, which shares the KV cache between layers to achieve layer-wise compression. Rather than intuitively sharing based on higher similarity, we discover a counterintuitive phenomenon: sharing dissimilar KV caches better preserves the model performance. Experiments show that KVSharer can reduce KV cache computation by 30%, thereby lowering memory consumption without significantly impacting model performance and it can also achieve at least 1.3 times generation acceleration. Although the KV cache Figure 1: Previous methods primarily focus greatly helps improve inference speed, it also significantly on discarding Keys and Values within layers. During the LLM inference In contrast, we share KV caches across layers phase, the KV cache typically accounts for based on their dissimilarity. Recent research has seen a proliferation of methods aimed at compressing KV caches to reduce memory consumption (Zandieh et al., 2024; Xu et al., 2024; Yang et al., 2024b; Zhang et al., 2024b;a; Dong et al., 2024). However, these efforts have predominantly focused on intra-layer KV cache compression within individual Transformer layers of LLM.

Videos are often used to learn or extract the necessary information to complete tasks in ways different than what text and static imagery alone can provide. However, many existing agent benchmarks neglect long-context video understanding, instead focusing on text or static image inputs. To bridge this gap, we introduce VideoWebArena (VideoWA), a benchmark for evaluating the capabilities of long-context multimodal agents for video understanding. VideoWA consists of 2,021 web agent tasks based on manually crafted video tutorials, which total almost four hours of content. For our benchmark, we define a taxonomy of long-context video-based agent tasks with two main areas of focus: skill retention and factual retention. While skill retention tasks evaluate whether an agent can use a given human demonstration to complete a task efficiently, the factual retention task evaluates whether an agent can retrieve instruction-relevant information from a video to complete a task. We find that the best model achieves 13.3% success on factual retention tasks and 45.8% on factual retention QA pairs, far below human performance at 73.9% and 79.3%, respectively. On skill retention tasks, long-context models perform worse with tutorials than without, exhibiting a 5% performance decrease in WebArena tasks and a 10.3% decrease in VisualWebArena tasks. Our work highlights the need to improve the agentic abilities of long-context multimodal models and provides a testbed for future development with long-context video agents.

Recent advancements in Large Language Models (LLMs) have demonstrated exceptional capabilities in natural language understanding and generation. While these models excel in general complex reasoning tasks, they still face challenges in mathematical problem-solving and logical reasoning. To address these limitations, researchers have explored function calling abilities, allowing LLMs to execute provided functions and utilize their outputs for task completion. However, concentrating on specific tasks can be very inefficient for large-scale LLMs to be used, because of the expensive cost of training and inference stages they need in terms of computational resources. This study introduces a novel framework for training smaller language models in function calling, focusing on specific logical and mathematical reasoning tasks. The approach aims to improve performances of small-scale models for these tasks using function calling, ensuring a high level of accuracy. Our framework employs an agent that, given a problem and a set of callable functions, queries the LLM by injecting a description and examples of the usable functions into the prompt and managing their calls in a step-by-step reasoning chain. This process is used to create a dataset of correct and incorrect reasoning chain chat completions from a large-scale LLM. This dataset is used to train a smaller LLM using Reinforcement Learning from Human Feedback (RLHF), specifically employing the Direct Preference Optimization (DPO) technique. Experimental results demonstrate how the proposed approach balances the trade-off between model size and performance, improving the ability of function calling for reasoning tasks, in smaller models.