Search

Conversational AI systems such as Alexa need to understand defective queries to ensure robust conversational understanding and reduce user friction. These defective queries often arise from user ambiguities, mistakes, or errors in automatic speech recognition (ASR) and natural language understanding (NLU). Personalized query rewriting is an approach that focuses on reducing defects in queries by taking into account the user's individual behavior and preferences. It typically relies on an index of past successful user interactions with the conversational AI. However, unseen interactions within the user's history present additional challenges for personalized query rewriting. This paper presents our "Collaborative Query Rewriting" approach, which specifically addresses the task of rewriting new user interactions that have not been previously observed in the user's history. This approach builds a "User Feedback Interaction Graph" (FIG) of historical user-entity interactions and leverages multi-hop graph traversal to enrich each user's index to cover future unseen defective queries. The enriched user index is called a Collaborative User Index and contains hundreds of additional entries. To counteract precision degradation from the enlarged index, we add additional transformer layers to the L1 retrieval model and incorporate graph-based and guardrail features into the L2 ranking model. Since the user index can be pre-computed, we further investigate the utilization of a Large Language Model (LLM) to enhance the FIG for user-entity link prediction in the Video/Music domains. Specifically, this paper investigates the Dolly-V2 7B model. We found that the user index augmented by the fine-tuned Dolly-V2 generation significantly enhanced the coverage of future unseen user interactions, thereby boosting QR performance on unseen queries compared with the graph traversal only approach.

Diffusion models (DMs) have recently been introduced in image deblurring and exhibited promising performance, particularly in terms of details reconstruction. However, the diffusion model requires a large number of inference iterations to recover the clean image from pure Gaussian noise, which consumes massive computational resources. Moreover, the distribution synthesized by the diffusion model is often misaligned with the target results, leading to restrictions in distortion-based metrics. To address the above issues, we propose the Hierarchical Integration Diffusion Model (HI-Diff), for realistic image deblurring. Specifically, we perform the DM in a highly compacted latent space to generate the prior feature for the deblurring process. The deblurring process is implemented by a regression-based method to obtain better distortion accuracy. Meanwhile, the highly compact latent space ensures the efficiency of the DM. Furthermore, we design the hierarchical integration module to fuse the prior into the regression-based model from multiple scales, enabling better generalization in complex blurry scenarios. Comprehensive experiments on synthetic and real-world blur datasets demonstrate that our HI-Diff outperforms state-of-the-art methods. Code and trained models are available at https://github.com/zhengchen1999/HI-Diff., Comment: Accepted to NeurIPS 2023 (Spotlight). Code is available at https://github.com/zhengchen1999/HI-Diff

Large language models (LLMs) have recently received significant attention for their exceptional capabilities. Despite extensive efforts in developing general-purpose LLMs that can be utilized in various natural language processing (NLP) tasks, there has been less research exploring their potential in recommender systems. In this paper, we propose a novel framework, named PALR, which aiming to combine user history behaviors (such as clicks, purchases, ratings, etc.) with LLMs to generate user preferred items. Specifically, we first use user/item interactions as guidance for candidate retrieval. Then we adopt a LLM-based ranking model to generate recommended items. Unlike existing approaches that typically adopt general-purpose LLMs for zero/few-shot recommendation testing or training on small-sized language models (with less than 1 billion parameters), which cannot fully elicit LLMs' reasoning abilities and leverage rich item side parametric knowledge, we fine-tune a 7 billion parameters LLM for the ranking purpose. This model takes retrieval candidates in natural language format as input, with instruction which explicitly asking to select results from input candidates during inference. Our experimental results demonstrate that our solution outperforms state-of-the-art models on various sequential recommendation tasks.

In this work, we focus on the communication aspect of decentralized learning, which involves multiple agents training a shared machine learning model using decentralized stochastic gradient descent (D-SGD) over distributed data. In particular, we investigate the impact of broadcast transmission and probabilistic random access policy on the convergence performance of D-SGD, considering the broadcast nature of wireless channels and the link dynamics in the communication topology. Our results demonstrate that optimizing the access probability to maximize the expected number of successful links is a highly effective strategy for accelerating the system convergence., Comment: 5 pages, 5 figures, accepted in IEEE SPAWC 2023

In this work, we consider a Federated Edge Learning (FEEL) system where training data are randomly generated over time at a set of distributed edge devices with long-term energy constraints. Due to limited communication resources and latency requirements, only a subset of devices is scheduled for participating in the local training process in every iteration. We formulate a stochastic network optimization problem for designing a dynamic scheduling policy that maximizes the time-average data importance from scheduled user sets subject to energy consumption and latency constraints. Our proposed algorithm based on the Lyapunov optimization framework outperforms alternative methods without considering time-varying data importance, especially when the generation of training data shows strong temporal correlation., Comment: Accepted for publication in the proceedings of 2023 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) workshop

Conventional document retrieval techniques are mainly based on the index-retrieve paradigm. It is challenging to optimize pipelines based on this paradigm in an end-to-end manner. As an alternative, generative retrieval represents documents as identifiers (docid) and retrieves documents by generating docids, enabling end-to-end modeling of document retrieval tasks. However, it is an open question how one should define the document identifiers. Current approaches to the task of defining document identifiers rely on fixed rule-based docids, such as the title of a document or the result of clustering BERT embeddings, which often fail to capture the complete semantic information of a document. We propose GenRet, a document tokenization learning method to address the challenge of defining document identifiers for generative retrieval. GenRet learns to tokenize documents into short discrete representations (i.e., docids) via a discrete auto-encoding approach. Three components are included in GenRet: (i) a tokenization model that produces docids for documents; (ii) a reconstruction model that learns to reconstruct a document based on a docid; and (iii) a sequence-to-sequence retrieval model that generates relevant document identifiers directly for a designated query. By using an auto-encoding framework, GenRet learns semantic docids in a fully end-to-end manner. We also develop a progressive training scheme to capture the autoregressive nature of docids and to stabilize training. We conduct experiments on the NQ320K, MS MARCO, and BEIR datasets to assess the effectiveness of GenRet. GenRet establishes the new state-of-the-art on the NQ320K dataset. Especially, compared to generative retrieval baselines, GenRet can achieve significant improvements on the unseen documents. GenRet also outperforms comparable baselines on MS MARCO and BEIR, demonstrating the method's generalizability.

We present a compact, long-term nearly continuous operation of a room-temperature Ca$^+$ optical clock setup towards a transportable clock, achieving an overall systematic uncertainty of $4.8\times 10^{-18}$ and an uptime rate of 97.8% over an 8-day period. The active liquid-cooling scheme is adopted, combined with the precise temperature measurement with 13 temperature sensors both inside and outside the vacuum chamber to ensure the accurate evaluation of the thermal environment for the optical clock. The environmental temperature uncertainty is evaluated as 293.31(0.4) K, corresponding to a blackbody radiation (BBR) frequency shift uncertainty of $4.6\times 10^{-18}$, which is reduced more than two times compared to our previous work. Through the frequency comparison between the room temperature Ca$^+$ optical clock and a cryogenic Ca$^+$ optical clock, the overall uncertainty of the clock comparison is $7.5\times 10^{-18}$, including a statistic uncertainty of $4.9\times 10^{-18}$ and a systematic uncertainty of $5.7\times 10^{-18}$. This work provides a set of feasible implementations for high-precision transportable ion optical clocks., Comment: 11 pages, 4 figures

Transformer architectures have exhibited remarkable performance in image super-resolution (SR). Since the quadratic computational complexity of the self-attention (SA) in Transformer, existing methods tend to adopt SA in a local region to reduce overheads. However, the local design restricts the global context exploitation, which is crucial for accurate image reconstruction. In this work, we propose the Recursive Generalization Transformer (RGT) for image SR, which can capture global spatial information and is suitable for high-resolution images. Specifically, we propose the recursive-generalization self-attention (RG-SA). It recursively aggregates input features into representative feature maps, and then utilizes cross-attention to extract global information. Meanwhile, the channel dimensions of attention matrices (query, key, and value) are further scaled to mitigate the redundancy in the channel domain. Furthermore, we combine the RG-SA with local self-attention to enhance the exploitation of the global context, and propose the hybrid adaptive integration (HAI) for module integration. The HAI allows the direct and effective fusion between features at different levels (local or global). Extensive experiments demonstrate that our RGT outperforms recent state-of-the-art methods quantitatively and qualitatively. Code and pre-trained models are available at https://github.com/zhengchen1999/RGT., Comment: Accepted to ICLR 2024. Code is available at https://github.com/zhengchen1999/RGT

Mixup-based data augmentation has been validated to be a critical stage in the self-training framework for unsupervised domain adaptive semantic segmentation (UDA-SS), which aims to transfer knowledge from a well-annotated (source) domain to an unlabeled (target) domain. Existing self-training methods usually adopt the popular region-based mixup techniques with a random sampling strategy, which unfortunately ignores the dynamic evolution of different semantics across various domains as training proceeds. To improve the UDA-SS performance, we propose an Informed Domain Adaptation (IDA) model, a self-training framework that mixes the data based on class-level segmentation performance, which aims to emphasize small-region semantics during mixup. In our IDA model, the class-level performance is tracked by an expected confidence score (ECS). We then use a dynamic schedule to determine the mixing ratio for data in different domains. Extensive experimental results reveal that our proposed method is able to outperform the state-of-the-art UDA-SS method by a margin of 1.1 mIoU in the adaptation of GTA-V to Cityscapes and of 0.9 mIoU in the adaptation of SYNTHIA to Cityscapes.

Prediction beyond partial observations is crucial for robots to navigate in unknown environments because it can provide extra information regarding the surroundings beyond the current sensing range or resolution. In this work, we consider the inpainting of semantic Bird's-Eye-View maps. We propose SePaint, an inpainting model for semantic data based on generative multinomial diffusion. To maintain semantic consistency, we need to condition the prediction for the missing regions on the known regions. We propose a novel and efficient condition strategy, Look-Back Condition (LB-Con), which performs one-step look-back operations during the reverse diffusion process. By doing so, we are able to strengthen the harmonization between unknown and known parts, leading to better completion performance. We have conducted extensive experiments on different datasets, showing our proposed model outperforms commonly used interpolation methods in various robotic applications.

Clinicians prescribe antibiotics by looking at the patient's health record with an experienced eye. However, the therapy might be rendered futile if the patient has drug resistance. Determining drug resistance requires time-consuming laboratory-level testing while applying clinicians' heuristics in an automated way is difficult due to the categorical or binary medical events that constitute health records. In this paper, we propose a novel framework for rapid clinical intervention by viewing health records as graphs whose nodes are mapped from medical events and edges as correspondence between events in given a time window. A novel graph-based model is then proposed to extract informative features and yield automated drug resistance analysis from those high-dimensional and scarce graphs. The proposed method integrates multi-task learning into a common feature extracting graph encoder for simultaneous analyses of multiple drugs as well as stabilizing learning. On a massive dataset comprising over 110,000 patients with urinary tract infections, we verify the proposed method is capable of attaining superior performance on the drug resistance prediction problem. Furthermore, automated drug recommendations resemblant to laboratory-level testing can also be made based on the model resistance analysis., Comment: 12 pages, 5 figures

Query Rewriting (QR) plays a critical role in large-scale dialogue systems for reducing frictions. When there is an entity error, it imposes extra challenges for a dialogue system to produce satisfactory responses. In this work, we propose KG-ECO: Knowledge Graph enhanced Entity COrrection for query rewriting, an entity correction system with corrupt entity span detection and entity retrieval/re-ranking functionalities. To boost the model performance, we incorporate Knowledge Graph (KG) to provide entity structural information (neighboring entities encoded by graph neural networks) and textual information (KG entity descriptions encoded by RoBERTa). Experimental results show that our approach yields a clear performance gain over two baselines: utterance level QR and entity correction without utilizing KG information. The proposed system is particularly effective for few-shot learning cases where target entities are rarely seen in training or there is a KG relation between the target entity and other contextual entities in the query.

We consider distributed average consensus in a wireless network with partial communication to reduce the number of transmissions in every iteration/round. Considering the broadcast nature of wireless channels, we propose a probabilistic approach that schedules a subset of nodes for broadcasting information to their neighbors in every round. We compare several heuristic methods for assigning the node broadcast probabilities under a fixed number of transmissions per round. Furthermore, we introduce a pre-compensation method to correct the bias between the consensus value and the average of the initial values, and suggest possible extensions for our design. Our results are particularly relevant for developing communication-efficient consensus protocols in a wireless environment with limited frequency/time resources., Comment: 5 pages, 3 figures, accepted for publication in IEEE Signal Processing Letters

Many policy optimization approaches in reinforcement learning incorporate a Kullback-Leilbler (KL) divergence to the previous policy, to prevent the policy from changing too quickly. This idea was initially proposed in a seminal paper on Conservative Policy Iteration, with approximations given by algorithms like TRPO and Munchausen Value Iteration (MVI). We continue this line of work by investigating a generalized KL divergence -- called the Tsallis KL divergence -- which use the $q$-logarithm in the definition. The approach is a strict generalization, as $q = 1$ corresponds to the standard KL divergence; $q > 1$ provides a range of new options. We characterize the types of policies learned under the Tsallis KL, and motivate when $q >1$ could be beneficial. To obtain a practical algorithm that incorporates Tsallis KL regularization, we extend MVI, which is one of the simplest approaches to incorporate KL regularization. We show that this generalized MVI($q$) obtains significant improvements over the standard MVI($q = 1$) across 35 Atari games., Comment: Accepted by NeurIPS 2023

Here we report the discovery of pressure-induced superconductivity in quasi-one-dimensional $\mathrm{Ta}_2 \mathrm{PdSe}_6$, through a combination of electrical transport, synchrotron x-ray diffraction, and theoretical calculations. Our transport measurements show that the superconductivity appears at a critical pressure $P_{\mathrm{c}} \sim 18.3$ GPa and is robust upon further compression up to $62.6$ GPa. The estimated upper critical field $\mu_0 H_{\mathrm{c} 2}(0)$ in the pressurized $\mathrm{Ta}_2 \mathrm{PdSe}_6$ is much lower than the Pauli limiting field, in contrast to the case in its isostructural analogs $M_2 \mathrm{Pd}_{\mathrm{x}} X_5$ $(M=\mathrm{Nb}$, Ta; $X=\mathrm{S}, \mathrm{Se})$. Concomitant with the occurrence of superconductivity, anomalies in pressuredependent transport properties are observed, including sign reversal of Hall coefficient, abnormally enhanced resistance, and dramatically suppressed magnetoresistance. Meanwhile, room-temperature synchrotron x-ray diffraction experiments reveal the stability of the pristine monoclinic structure (space group $C 2 / m$ ) upon compression. Combined with the density functional theory calculations, we argue that a pressure-induced Lifshitz transition could be the electronic origin of the emergent superconductivity in $\mathrm{Ta}_2 \mathrm{PdSe}_6$., Comment: 7 pages, 7 figures

There has been increasing interest in investigating the possible parity violating features in the gravity theory and on the cosmological scales. In this work, we consider a class of scalar-nonmetricity theory, of which the Lagrangian is polynomial built of the nonmetricity tensor and a scalar field. The nonmetricity tensor is coupled with the scalar field through its first order derivative. Besides the monomials that are quadratic order in the nonmetricity tensor, we also construct monomials that are cubic order in the nonmetricity tensor in both the parity preserving and violating cases. These monomials act as the non-canonical (i.e., non-quadratic) kinetic terms for the spacetime metric, and will change the behavior in the propagation of the gravitational waves. We find that the gravitational waves are generally polarized, which present both the amplitude and velocity birefringence features due to the parity violation of the theory. Due to the term proportional to $1/k$ in the phase velocities, one of the two polarization modes suffers from the gradient instability on large scales., Comment: 12 pages, no figure

Federated Learning (FL) is a collaborative machine learning (ML) framework that combines on-device training and server-based aggregation to train a common ML model among distributed agents. In this work, we propose an asynchronous FL design with periodic aggregation to tackle the straggler issue in FL systems. Considering limited wireless communication resources, we investigate the effect of different scheduling policies and aggregation designs on the convergence performance. Driven by the importance of reducing the bias and variance of the aggregated model updates, we propose a scheduling policy that jointly considers the channel quality and training data representation of user devices. The effectiveness of our channel-aware data-importance-based scheduling policy, compared with state-of-the-art methods proposed for synchronous FL, is validated through simulations. Moreover, we show that an ``age-aware'' aggregation weighting design can significantly improve the learning performance in an asynchronous FL setting.

Motivated by real-world applications, we study the fair allocation of graphical resources, where the resources are the vertices in a graph. Upon receiving a set of resources, an agent's utility equals the weight of a maximum matching in the induced subgraph. We care about maximin share (MMS) fairness and envy-freeness up to one item (EF1). Regarding MMS fairness, the problem does not admit a finite approximation ratio for heterogeneous agents. For homogeneous agents, we design constant-approximation polynomial-time algorithms, and also note that significant amount of social welfare is sacrificed inevitably in order to ensure (approximate) MMS fairness. We then consider EF1 allocations whose existence is guaranteed. However, the social welfare guarantee of EF1 allocations cannot be better than $1/n$ for the general case, where $n$ is the number of agents.Fortunately, for three special cases, binary-weight, two-agents and homogeneous-agents, we are able to design polynomial-time algorithms that also ensure a constant fractions of the maximum social welfare.