Search

Increasingly powerful Large Language Model (LLM) based chatbots, like ChatGPT and Bard, are becoming available to users that have the potential to revolutionize the quality of decision-making achieved by the public. In this context, we set out to investigate how such systems perform in the personal finance domain, where financial inclusion has been an overarching stated aim of banks for decades. We asked 13 questions representing banking products in personal finance: bank account, credit card, and certificate of deposits and their inter-product interactions, and decisions related to high-value purchases, payment of bank dues, and investment advice, and in different dialects and languages (English, African American Vernacular English, and Telugu). We find that although the outputs of the chatbots are fluent and plausible, there are still critical gaps in providing accurate and reliable financial information using LLM-based chatbots.

Game level blending via machine learning, the process of combining features of game levels to create unique and novel game levels using Procedural Content Generation via Machine Learning (PCGML) techniques, has gained increasing popularity in recent years. However, many existing techniques rely on human-annotated level representations, which limits game level blending to a limited number of annotated games. Even with annotated games, researchers often need to author an additional shared representation to make blending possible. In this paper, we present a novel approach to game level blending that employs Clustering-based Tile Embeddings (CTE), a learned level representation technique that can serve as a level representation for unannotated games and a unified level representation across games without the need for human annotation. CTE represents game level tiles as a continuous vector representation, unifying their visual, contextual, and behavioral information. We apply this approach to two classic Nintendo games, Lode Runner and The Legend of Zelda. We run an evaluation comparing the CTE representation to a common, human-annotated representation in the blending task and find that CTE has comparable or better performance without the need for human annotation., Comment: 8 pages, 3 figures

Existing reference-free turn-level evaluation metrics for chatbots inadequately capture the interaction between the user and the system. Consequently, they often correlate poorly with human evaluations. To address this issue, we propose a novel model-agnostic approach that leverages Conditional Pointwise Mutual Information (C-PMI) to measure the turn-level interaction between the system and the user based on a given evaluation dimension. Experimental results on the widely used FED dialogue evaluation dataset demonstrate that our approach significantly improves the correlation with human judgment compared with existing evaluation systems. By replacing the negative log-likelihood-based scorer with our proposed C-PMI scorer, we achieve a relative 62.6% higher Spearman correlation on average for the FED evaluation metric. Our code is publicly available at https://github.com/renll/C-PMI., Comment: Published at ACL2023 DialDoc Workshop; Updated Results

Retrieve-and-rerank is a prevalent framework in neural information retrieval, wherein a bi-encoder network initially retrieves a pre-defined number of candidates (e.g., K=100), which are then reranked by a more powerful cross-encoder model. While the reranker often yields improved candidate scores compared to the retriever, its scope is confined to only the top K retrieved candidates. As a result, the reranker cannot improve retrieval performance in terms of Recall@K. In this work, we propose to leverage the reranker to improve recall by making it provide relevance feedback to the retriever at inference time. Specifically, given a test instance during inference, we distill the reranker's predictions for that instance into the retriever's query representation using a lightweight update mechanism. The aim of the distillation loss is to align the retriever's candidate scores more closely with those produced by the reranker. The algorithm then proceeds by executing a second retrieval step using the updated query vector. We empirically demonstrate that this method, applicable to various retrieve-and-rerank frameworks, substantially enhances retrieval recall across multiple domains, languages, and modalities., Comment: Preprint

In this work, we investigate the device-to-device variations in remanent polarization of Hafnium-Zirconium-Oxide based Metal-Ferroelectric-Insulator-Metal (MFIM) stacks. We consider the effects of polycrystallinity in conjunction with multi-domain effects in HZO to understand the dependencies of variations on static and dynamic voltage stimuli using our 3D dynamic multi-grain phase-field simulation framework. We examine the trends in variations due to various design factors - set voltage, pulse amplitude and pulse width and correlate them to the dynamics of polarization switching and the underlying mechanisms. According to our analysis, variations exhibit a non-monotonic dependence on set voltage due to the interplay between voltage-dependent switching mechanisms and the polycrystalline structure. We further report that towards the higher end of the set voltages, collapsing of oppositely polarized domains can lead to increase in variations. We also show that ferroelectric thickness scaling lowers the device-to-device variations. In addition, considering the dynamics of polarization switching, we signify the key role of voltage and temporal dependence of domain nucleation in dictating the trends in variations. Finally, we show that to reach a target mean polarization, using a pulse with lower amplitude for longer duration results in lower variations compared to higher amplitude pulse for a shorter duration.

Timely and comprehensive understanding of emerging events is crucial for effective decision-making; automating situation report generation can significantly reduce the time, effort, and cost for intelligence analysts. In this work, we identify intelligence analysts' practices and preferences for AI assistance in situation report generation to guide the design strategies for an effective, trust-building interface that aligns with their thought processes and needs. Next, we introduce SmartBook, an automated framework designed to generate situation reports from large volumes of news data, creating structured reports by automatically discovering event-related strategic questions. These reports include multiple hypotheses (claims), summarized and grounded to sources with factual evidence, to promote in-depth situation understanding. Our comprehensive evaluation of SmartBook, encompassing a user study alongside a content review with an editing study, reveals SmartBook's effectiveness in generating accurate and relevant situation reports. Qualitative evaluations indicate over 80% of questions probe for strategic information, and over 90% of summaries produce tactically useful content, being consistently favored over summaries from a large language model integrated with web search. The editing study reveals that minimal information is removed from the generated text (under 2.5%), suggesting that SmartBook provides analysts with a valuable foundation for situation reports, Comment: Preprint

In this paper, a new demonstration-based path-planning framework for the visual inspection of large structures using UAVs is proposed. We introduce UPPLIED: UAV Path PLanning for InspEction through Demonstration, which utilizes a demonstrated trajectory to generate a new trajectory to inspect other structures of the same kind. The demonstrated trajectory can inspect specific regions of the structure and the new trajectory generated by UPPLIED inspects similar regions in the other structure. The proposed method generates inspection points from the demonstrated trajectory and uses standardization to translate those inspection points to inspect the new structure. Finally, the position of these inspection points is optimized to refine their view. Numerous experiments were conducted with various structures and the proposed framework was able to generate inspection trajectories of various kinds for different structures based on the demonstration. The trajectories generated match with the demonstrated trajectory in geometry and at the same time inspect the regions inspected by the demonstration trajectory with minimum deviation. The experimental video of the work can be found at https://youtu.be/YqPx-cLkv04., Comment: Accepted for publication in IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS 2023), Detroit, Michigan, USA

A primary objective of news articles is to establish the factual record for an event, frequently achieved by conveying both the details of the specified event (i.e., the 5 Ws; Who, What, Where, When and Why regarding the event) and how people reacted to it (i.e., reported statements). However, existing work on news summarization almost exclusively focuses on the event details. In this work, we propose the novel task of summarizing the reactions of different speakers, as expressed by their reported statements, to a given event. To this end, we create a new multi-document summarization benchmark, SUMREN, comprising 745 summaries of reported statements from various public figures obtained from 633 news articles discussing 132 events. We propose an automatic silver training data generation approach for our task, which helps smaller models like BART achieve GPT-3 level performance on this task. Finally, we introduce a pipeline-based framework for summarizing reported speech, which we empirically show to generate summaries that are more abstractive and factual than baseline query-focused summarization approaches., Comment: Accepted at AAAI 2023

This paper presents MOCAS, a multimodal dataset dedicated for human cognitive workload (CWL) assessment. In contrast to existing datasets based on virtual game stimuli, the data in MOCAS was collected from realistic closed-circuit television (CCTV) monitoring tasks, increasing its applicability for real-world scenarios. To build MOCAS, two off-the-shelf wearable sensors and one webcam were utilized to collect physiological signals and behavioral features from 21 human subjects. After each task, participants reported their CWL by completing the NASA-Task Load Index (NASA-TLX) and Instantaneous Self-Assessment (ISA). Personal background (e.g., personality and prior experience) was surveyed using demographic and Big Five Factor personality questionnaires, and two domains of subjective emotion information (i.e., arousal and valence) were obtained from the Self-Assessment Manikin (SAM), which could serve as potential indicators for improving CWL recognition performance. Technical validation was conducted to demonstrate that target CWL levels were elicited during simultaneous CCTV monitoring tasks; its results support the high quality of the collected multimodal signals., Comment: 18 pages, 10 figures, 9 tables, Accepted IEEE transaction Affective Computing

Twisted string actuators (TSAs) have exhibited great promise in robotic applications by generating high translational force with low input torque. To further facilitate their robotic applications, it is strongly desirable but challenging to enhance their consistent strain generation while maintaining compliance. Existing studies predominantly considered overtwisting and coiling after the regular twisting stage to be undesirable non-uniform and unpredictable knots, entanglements, and coils formed to create an unstable and failure-prone structure. Overtwisting would work well for TSAs when uniform coils can be consistently formed. In this study, we realize uniform and consistent coil formation in overtwisted TSAs, which greatly increases their strain. Furthermore, we investigate methods for enabling uniform coil formation upon overtwisting the strings in a TSA and present a procedure to systematically "train" the strings. To the authors' best knowledge, this is the first study to experimentally investigate overtwisting for TSAs with different stiffnesses and realize consistent uniform coil formation. Ultra-high molecular-weight polyethylene (UHMWPE) strings form the stiff TSAs whereas compliant TSAs are realized with stretchable and conductive supercoiled polymer (SCP) strings. The strain, force, velocity, and torque of each overtwisted TSA was studied. Overtwisting and coiling resulted in approximately 70% strain in stiff TSAs and approximately 60% strain in compliant TSAs. This is more than twice the strain achieved through regular twisting. Lastly, the overtwisted TSA was successfully demonstrated in a robotic bicep.

Realizing high-performance soft robotic grippers is challenging because of the inherent limitations of the soft actuators and artificial muscles that drive them, including low force output, small actuation range, and poor compactness. Despite advances in this area, realizing compact soft grippers with high dexterity and force output is still challenging. This paper explores twisted string actuators (TSAs) to drive a soft robotic gripper. TSAs have been used in numerous robotic applications, but their inclusion in soft robots has been limited. The proposed design of the gripper was inspired by the human hand. Tunable stiffness was implemented in the fingers with antagonistic TSAs. The fingers' bending angles, actuation speed, blocked force output, and stiffness tuning were experimentally characterized. The gripper achieved a score of 6 on the Kapandji test and recreated 31 of the 33 grasps of the Feix GRASP taxonomy. It exhibited a maximum grasping force of 72 N, which was almost 13 times its own weight. A comparison study revealed that the proposed gripper exhibited equivalent or superior performance compared to other similar soft grippers., Comment: 19 pages, 15 figures