Offline Multi-agent Reinforcement Learning (MARL) is valuable in scenarios where online interaction is impractical or risky. While independent learning in MARL offers flexibility and scalability, accurately assigning credit to individual agents in offline settings poses challenges due to partial observability and emergent behavior. Directly transferring the online credit assignment method to offline settings results in suboptimal outcomes due to the absence of real-time feedback and intricate agent interactions. Our approach, MACCA, characterizing the generative process as a Dynamic Bayesian Network, captures relationships between environmental variables, states, actions, and rewards. Estimating this model on offline data, MACCA can learn each agent's contribution by analyzing the causal relationship of their individual rewards, ensuring accurate and interpretable credit assignment. Additionally, the modularity of our approach allows it to seamlessly integrate with various offline MARL methods. Theoretically, we proved that under the setting of the offline dataset, the underlying causal structure and the function for generating the individual rewards of agents are identifiable, which laid the foundation for the correctness of our modeling. Experimentally, we tested MACCA in two environments, including discrete and continuous action settings. The results show that MACCA outperforms SOTA methods and improves performance upon their backbones.
Offline goal-conditioned RL (GCRL) offers a feasible paradigm to learn general-purpose policies from diverse and multi-task offline datasets. Despite notable recent progress, the predominant offline GCRL methods have been restricted to model-free approaches, constraining their capacity to tackle limited data budgets and unseen goal generalization. In this work, we propose a novel two-stage model-based framework, Goal-conditioned Offline Planning (GOPlan), including (1) pretraining a prior policy capable of capturing multi-modal action distribution within the multi-goal dataset; (2) employing the reanalysis method with planning to generate imagined trajectories for funetuning policies. Specifically, the prior policy is based on an advantage-weighted Conditioned Generative Adversarial Networks that exhibits distinct mode separation to overcome the pitfalls of out-of-distribution (OOD) actions. For further policy optimization, the reanalysis method generates high-quality imaginary data by planning with learned models for both intra-trajectory and inter-trajectory goals. Through experimental evaluations, we demonstrate that GOPlan achieves state-of-the-art performance on various offline multi-goal manipulation tasks. Moreover, our results highlight the superior ability of GOPlan to handle small data budgets and generalize to OOD goals.
Dialogue state tracking (DST) plays an important role in task-oriented dialogue systems. However, collecting a large amount of turn-by-turn annotated dialogue data is costly and inefficient. In this paper, we propose a novel turn-level active learning framework for DST to actively select turns in dialogues to annotate. Given the limited labelling budget, experimental results demonstrate the effectiveness of selective annotation of dialogue turns. Additionally, our approach can effectively achieve comparable DST performance to traditional training approaches with significantly less annotated data, which provides a more efficient way to annotate new dialogue data.
Continual learning (CL) is a paradigm that aims to replicate the human ability to learn and accumulate knowledge continually without forgetting previous knowledge and transferring it to new tasks. Recent instruction tuning (IT) involves fine-tuning models to make them more adaptable to solving NLP tasks in general. However, it is still uncertain how instruction tuning works in the context of CL tasks. This challenging yet practical problem is formulated as Continual Instruction Tuning (CIT). In this work, we establish a CIT benchmark consisting of learning and evaluation protocols. We curate two long dialogue task streams of different types, InstrDialog and InstrDialog++, to study various CL methods systematically. Our experiments show that existing CL methods do not effectively leverage the rich natural language instructions, and fine-tuning an instruction-tuned model sequentially can yield similar or better results. We further explore different aspects that might affect the learning of CIT. We hope this benchmark will facilitate more research in this direction.
Query rewriting plays a vital role in enhancing conversational search by transforming context-dependent user queries into standalone forms. Existing approaches primarily leverage human-rewritten queries as labels to train query rewriting models. However, human rewrites may lack sufficient information for optimal retrieval performance. To overcome this limitation, we propose utilizing large language models (LLMs) as query rewriters, enabling the generation of informative query rewrites through well-designed instructions. We define four essential properties for well-formed rewrites and incorporate all of them into the instruction. In addition, we introduce the role of rewrite editors for LLMs when initial query rewrites are available, forming a "rewrite-then-edit" process. Furthermore, we propose distilling the rewriting capabilities of LLMs into smaller models to reduce rewriting latency. Our experimental evaluation on the QReCC dataset demonstrates that informative query rewrites can yield substantially improved retrieval performance compared to human rewrites, especially with sparse retrievers.
Although large language models (LLMs) are impressive in solving various tasks, they can quickly be outdated after deployment. Maintaining their up-to-date status is a pressing concern in the current era. This paper provides a comprehensive review of recent advances in aligning LLMs with the ever-changing world knowledge without re-training from scratch. We categorize research works systemically and provide in-depth comparisons and discussion. We also discuss existing challenges and highlight future directions to facilitate research in this field. We release the paper list at https://github.com/hyintell/awesome-refreshing-llms
Data-centric AI, with its primary focus on the collection, management, and utilization of data to drive AI models and applications, has attracted increasing attention in recent years. In this article, we conduct an in-depth and comprehensive review, offering a forward-looking outlook on the current efforts in data-centric AI pertaining to graph data-the fundamental data structure for representing and capturing intricate dependencies among massive and diverse real-life entities. We introduce a systematic framework, Data-centric Graph Machine Learning (DC-GML), that encompasses all stages of the graph data lifecycle, including graph data collection, exploration, improvement, exploitation, and maintenance. A thorough taxonomy of each stage is presented to answer three critical graph-centric questions: (1) how to enhance graph data availability and quality; (2) how to learn from graph data with limited-availability and low-quality; (3) how to build graph MLOps systems from the graph data-centric view. Lastly, we pinpoint the future prospects of the DC-GML domain, providing insights to navigate its advancements and applications.
Accurate human mobility prediction underpins many important applications across a variety of domains, including epidemic modelling, transport planning, and emergency responses. Due to the sparsity of mobility data and the stochastic nature of people's daily activities, achieving precise predictions of people's locations remains a challenge. While recently developed large language models (LLMs) have demonstrated superior performance across numerous language-related tasks, their applicability to human mobility studies remains unexplored. Addressing this gap, this article delves into the potential of LLMs for human mobility prediction tasks. We introduce a novel method, LLM-Mob, which leverages the language understanding and reasoning capabilities of LLMs for analysing human mobility data. We present concepts of historical stays and context stays to capture both long-term and short-term dependencies in human movement and enable time-aware prediction by using time information of the prediction target. Additionally, we design context-inclusive prompts that enable LLMs to generate more accurate predictions. Comprehensive evaluations of our method reveal that LLM-Mob excels in providing accurate and interpretable predictions, highlighting the untapped potential of LLMs in advancing human mobility prediction techniques. We posit that our research marks a significant paradigm shift in human mobility modelling, transitioning from building complex domain-specific models to harnessing general-purpose LLMs that yield accurate predictions through language instructions. The code for this work is available at https://github.com/xlwang233/LLM-Mob.
Despite the fact that adversarial training has become the de facto method for improving the robustness of deep neural networks, it is well-known that vanilla adversarial training suffers from daunting robust overfitting, resulting in unsatisfactory robust generalization. A number of approaches have been proposed to address these drawbacks such as extra regularization, adversarial weights perturbation, and training with more data over the last few years. However, the robust generalization improvement is yet far from satisfactory. In this paper, we approach this challenge with a brand new perspective -- refining historical optimization trajectories. We propose a new method named \textbf{Weighted Optimization Trajectories (WOT)} that leverages the optimization trajectories of adversarial training in time. We have conducted extensive experiments to demonstrate the effectiveness of WOT under various state-of-the-art adversarial attacks. Our results show that WOT integrates seamlessly with the existing adversarial training methods and consistently overcomes the robust overfitting issue, resulting in better adversarial robustness. For example, WOT boosts the robust accuracy of AT-PGD under AA-$L_{\infty}$ attack by 1.53\% $\sim$ 6.11\% and meanwhile increases the clean accuracy by 0.55\%$\sim$5.47\% across SVHN, CIFAR-10, CIFAR-100, and Tiny-ImageNet datasets.
We propose a novel value approximation method, namely Eigensubspace Regularized Critic (ERC) for deep reinforcement learning (RL). ERC is motivated by an analysis of the dynamics of Q-value approximation error in the Temporal-Difference (TD) method, which follows a path defined by the 1-eigensubspace of the transition kernel associated with the Markov Decision Process (MDP). It reveals a fundamental property of TD learning that has remained unused in previous deep RL approaches. In ERC, we propose a regularizer that guides the approximation error tending towards the 1-eigensubspace, resulting in a more efficient and stable path of value approximation. Moreover, we theoretically prove the convergence of the ERC method. Besides, theoretical analysis and experiments demonstrate that ERC effectively reduces the variance of value functions. Among 26 tasks in the DMControl benchmark, ERC outperforms state-of-the-art methods for 20. Besides, it shows significant advantages in Q-value approximation and variance reduction. Our code is available at https://sites.google.com/view/erc-ecml23/.