基于会话的推荐系统（SBRS）表现出优于常规方法的性能。但是，它们在大规模工业数据集上显示出有限的可伸缩性，因为大多数模型都会学习一个嵌入每个项目。这导致了巨大的记忆要求（每项存储一个矢量），并且在稀疏的会话上具有冷启动或不受欢迎的项目的性能差。使用一个公共和一个大型工业数据集，我们在实验上表明，最先进的SBRS在稀疏项目的稀疏会议上的性能较低。我们提出了M2TREC，这是一种基于会话建议的元数据感知的多任务变压器模型。我们提出的方法学习了从项目元数据到嵌入的转换函数，因此是免费的（即，不需要学习一个嵌入每个项目）。它集成了项目元数据以学习各种项目属性的共享表示。在推论期间，将为与先前在培训期间观察到的项目共享的属性分配新的或不受欢迎的项目，因此将与这些项目具有相似的表示，从而使甚至冷启动和稀疏项目的建议。此外，M2TREC接受了多任务设置的培训，以预测会话中的下一个项目及其主要类别和子类别。我们的多任务策略使该模型收敛更快，并显着改善了整体性能。实验结果表明，使用我们在两个数据集中稀疏项目上提出的方法进行了显着的性能增长。

基于会话的建议系统在会话中捕获用户的短期兴趣。会话上下文（即，会话中用户在会话中的高级兴趣或意图）在大多数数据集中都没有明确给出，并且隐式推断会话上下文作为项目级属性的汇总是粗略的。在本文中，我们提出了ISCON，该ISCON隐含地将会议上下文化。ISCON首先通过创建会话信息图，学习图嵌入和聚类来为会话生成隐式上下文，以将会话分配给上下文。然后，ISCON训练会话上下文预测器，并使用预测上下文的嵌入来增强下一项目的预测准确性。四个数据集的实验表明，ISCON比最新模型具有优越的下一项目预测准确性。REDDIT数据集中的ISCON的案例研究证实，分配的会话上下文是独特而有意义的。

预测模型可以表现出对培训数据的敏感性：训练数据中的微小变化可以产生在测试时间期间为单个数据点分配相互矛盾的预测的模型。在这项工作中，我们研究了推荐系统中的这种敏感性，其中用户的建议在其他无关用户的交互中受到较小的扰动的巨大改变。我们介绍了推荐系统的稳定性度量，称为等级列表灵敏度（RLS），该量度衡量了由于培训数据中的扰动而导致的测试时间变化时在测试时间变化时如何生成的等级列表。我们开发了一种方法，即Casper，该方法使用级联效应来识别最小和系统的扰动，以在推荐系统中诱导更高的不稳定性。四个数据集的实验表明，推荐模型对引入或通过Casper引入的次要扰动过于敏感 - 甚至将一个用户的一个随机交互扰动会大大更改所有用户的建议列表。重要的是，借助Casper扰动，这些模型比高准确性的使用者（即那些接受低质量建议的人）为低临界用户（即那些接受低质量建议的人）产生更多的不稳定建议。

Generative AI has matured to a point where large-scale models can generate text that seems indistinguishable from human-written text and remarkably photorealistic images. Automatically measuring how close the distribution of generated data is to the target real data distribution is a key step in diagnosing existing models and developing better models. We present MAUVE, a family of comparison measures between pairs of distributions such as those encountered in the generative modeling of text or images. These scores are statistical summaries of divergence frontiers capturing two types of errors in generative modeling. We explore four approaches to statistically estimate these scores: vector quantization, non-parametric estimation, classifier-based estimation, and parametric Gaussian approximations. We provide statistical bounds for the vector quantization approach. Empirically, we find that the proposed scores paired with a range of $f$-divergences and statistical estimation methods can quantify the gaps between the distributions of human-written text and those of modern neural language models by correlating with human judgments and identifying known properties of the generated texts. We conclude the paper by demonstrating its applications to other AI domains and discussing practical recommendations.

We study model-based reinforcement learning (RL) for episodic Markov decision processes (MDP) whose transition probability is parametrized by an unknown transition core with features of state and action. Despite much recent progress in analyzing algorithms in the linear MDP setting, the understanding of more general transition models is very restrictive. In this paper, we establish a provably efficient RL algorithm for the MDP whose state transition is given by a multinomial logistic model. To balance the exploration-exploitation trade-off, we propose an upper confidence bound-based algorithm. We show that our proposed algorithm achieves $\tilde{\mathcal{O}}(d \sqrt{H^3 T})$ regret bound where $d$ is the dimension of the transition core, $H$ is the horizon, and $T$ is the total number of steps. To the best of our knowledge, this is the first model-based RL algorithm with multinomial logistic function approximation with provable guarantees. We also comprehensively evaluate our proposed algorithm numerically and show that it consistently outperforms the existing methods, hence achieving both provable efficiency and practical superior performance.

This work presents a detailed linguistic analysis into why larger Transformer-based pre-trained language models with more parameters and lower perplexity nonetheless yield surprisal estimates that are less predictive of human reading times. First, regression analyses show a strictly monotonic, positive log-linear relationship between perplexity and fit to reading times for the more recently released five GPT-Neo variants and eight OPT variants on two separate datasets, replicating earlier results limited to just GPT-2 (Oh et al., 2022). Subsequently, analysis of residual errors reveals a systematic deviation of the larger variants, such as underpredicting reading times of named entities and making compensatory overpredictions for reading times of function words such as modals and conjunctions. These results suggest that the propensity of larger Transformer-based models to 'memorize' sequences during training makes their surprisal estimates diverge from humanlike expectations, which warrants caution in using pre-trained language models to study human language processing.

Generalisation to unseen contexts remains a challenge for embodied navigation agents. In the context of semantic audio-visual navigation (SAVi) tasks, the notion of generalisation should include both generalising to unseen indoor visual scenes as well as generalising to unheard sounding objects. However, previous SAVi task definitions do not include evaluation conditions on truly novel sounding objects, resorting instead to evaluating agents on unheard sound clips of known objects; meanwhile, previous SAVi methods do not include explicit mechanisms for incorporating domain knowledge about object and region semantics. These weaknesses limit the development and assessment of models' abilities to generalise their learned experience. In this work, we introduce the use of knowledge-driven scene priors in the semantic audio-visual embodied navigation task: we combine semantic information from our novel knowledge graph that encodes object-region relations, spatial knowledge from dual Graph Encoder Networks, and background knowledge from a series of pre-training tasks -- all within a reinforcement learning framework for audio-visual navigation. We also define a new audio-visual navigation sub-task, where agents are evaluated on novel sounding objects, as opposed to unheard clips of known objects. We show improvements over strong baselines in generalisation to unseen regions and novel sounding objects, within the Habitat-Matterport3D simulation environment, under the SoundSpaces task.

Transformer-based large language models are trained to make predictions about the next word by aggregating representations of previous tokens through their self-attention mechanism. In the field of cognitive modeling, such attention patterns have recently been interpreted as embodying the process of cue-based retrieval, in which attention over multiple targets is taken to generate interference and latency during retrieval. Under this framework, this work first defines an entropy-based predictor that quantifies the diffuseness of self-attention, as well as distance-based predictors that capture the incremental change in attention patterns across timesteps. Moreover, following recent studies that question the informativeness of attention weights, we also experiment with alternative methods for incorporating vector norms into attention weights. Regression experiments using predictors calculated from the GPT-2 language model show that these predictors deliver a substantially better fit to held-out self-paced reading and eye-tracking data over a rigorous baseline including GPT-2 surprisal. Additionally, the distance-based predictors generally demonstrated higher predictive power, with effect sizes of up to 6.59 ms per standard deviation on self-paced reading times (compared to 2.82 ms for surprisal) and 1.05 ms per standard deviation on eye-gaze durations (compared to 3.81 ms for surprisal).

Task-oriented dialogue (TOD) systems are mainly based on the slot-filling-based TOD (SF-TOD) framework, in which dialogues are broken down into smaller, controllable units (i.e., slots) to fulfill a specific task. A series of approaches based on this framework achieved remarkable success on various TOD benchmarks. However, we argue that the current TOD benchmarks are limited to surrogate real-world scenarios and that the current TOD models are still a long way from unraveling the scenarios. In this position paper, we first identify current status and limitations of SF-TOD systems. After that, we explore the WebTOD framework, the alternative direction for building a scalable TOD system when a web/mobile interface is available. In WebTOD, the dialogue system learns how to understand the web/mobile interface that the human agent interacts with, powered by a large-scale language model.

We outline our work on evaluating robots that assist older adults by engaging with them through multiple modalities that include physical interaction. Our thesis is that to increase the effectiveness of assistive robots: 1) robots need to understand and effect multimodal actions, 2) robots should not only react to the human, they need to take the initiative and lead the task when it is necessary. We start by briefly introducing our proposed framework for multimodal interaction and then describe two different experiments with the actual robots. In the first experiment, a Baxter robot helps a human find and locate an object using the Multimodal Interaction Manager (MIM) framework. In the second experiment, a NAO robot is used in the same task, however, the roles of the robot and the human are reversed. We discuss the evaluation methods that were used in these experiments, including different metrics employed to characterize the performance of the robot in each case. We conclude by providing our perspective on the challenges and opportunities for the evaluation of assistive robots for older adults in realistic settings.