Reasoning in Large Language Models (LLMs) has recently shown strong potential in enhancing generative recommendation through deep understanding of complex user preference. Existing approaches follow a reason-then-recommend paradigm, where LLMs perform step-by-step reasoning before item generation. However, this paradigm inevitably suffers from reasoning degradation (i.e., homogeneous or error-accumulated reasoning) due to the lack of intermediate verification, thus undermining the recommendation. To bridge this gap, we propose a novel reason-verify-recommend paradigm, which interleaves reasoning with verification to provide reliable feedback, guiding the reasoning process toward more faithful user preference understanding. To enable effective verification, we establish two key principles for verifier design: 1) reliability ensures accurate evaluation of reasoning correctness and informative guidance generation; and 2) multi-dimensionality emphasizes comprehensive verification across multi-dimensional user preferences. Accordingly, we propose an effective implementation called VRec. It employs a mixture of verifiers to ensure multi-dimensionality, while leveraging a proxy prediction objective to pursue reliability. Experiments on four real-world datasets demonstrate that VRec substantially enhances recommendation effectiveness and scalability without compromising efficiency.

The task of item-to-item (I2I) retrieval is to identify a set of relevant and highly engaging items based on a given trigger item. It is a crucial component in modern recommendation systems, where users' previously engaged items serve as trigger items to retrieve relevant content for future engagement. However, existing I2I retrieval models in industry are primarily built on co-engagement data and optimized using the recall measure, which overly emphasizes co-engagement patterns while failing to capture semantic relevance. This often leads to overfitting short-term co-engagement trends at the expense of long-term benefits such as discovering novel interests and promoting content diversity. To address this challenge, we propose MTMH, a Multi-Task and Multi-Head I2I retrieval model that achieves both high recall and semantic relevance. Our model consists of two key components: 1) a multi-task learning loss for formally optimizing the trade-off between recall and semantic relevance, and 2) a multi-head I2I retrieval architecture for retrieving both highly co-engaged and semantically relevant items. We evaluate MTMH using proprietary data from a commercial platform serving billions of users and demonstrate that it can improve recall by up to 14.4% and semantic relevance by up to 56.6% compared with prior state-of-the-art models. We also conduct live experiments to verify that MTMH can enhance both short-term consumption metrics and long-term user-experience-related metrics. Our work provides a principled approach for jointly optimizing I2I recall and semantic relevance, which has significant implications for improving the overall performance of recommendation systems.

Specialized machine learning (ML) models tailored to users’ needs and requests are increasingly being deployed on smart devices with cameras, to provide personalized intelligent services taking advantage of camera data. However, two primary challenges hinder the training of such models: the lack of publicly available labeled data suitable for specialized tasks and the inaccessibility of labeled private data due to concerns about user privacy. To address these challenges, we propose a novel system SpinML, where the server generates customized Synthetic image data to Privately traIN a specialized ML model tailored to the user request, with the usage of only a few sanitized reference images from the user. SpinML offers users fine-grained, object-level control over the reference images, which allows user to trade between the privacy and utility of the generated synthetic data according to their privacy preferences. Through experiments on three specialized model training tasks, we demonstrate that our proposed system can enhance the perfor- mance of specialized models without compromising users’ privacy preferences.

Toxic content detection is crucial for online services to remove inappropriate content that violates community standards. To automate the detection process, prior works have proposed varieties of machine learning (ML) approaches to train Language Models (LMs) for toxic content detection. However, both their accuracy and transferability across datasets are limited. Recently, Large Language Models (LLMs) have shown promise in toxic content detection due to their superior zero-shot and few-shot in-context learning ability as well as broad transferability on ML tasks. However, efficiently designing prompts for LLMs remains challenging. Moreover, the high run-time cost of LLMs may hinder their deployments in production. To address these challenges, in this work, we propose BD-LLM, a novel and efficient approach to bootstrapping and distilling LLMs for toxic content detection. Specifically, we design a novel prompting method named Decision-Tree-of-Thought (DToT) to bootstrap LLMs' detection performance and extract high-quality rationales. DToT can automatically select more fine-grained context to re-prompt LLMs when their responses lack confidence. Additionally, we use the rationales extracted via DToT to fine-tune student LMs. Our experimental results on various datasets demonstrate that DToT can improve the accuracy of LLMs by up to 4.6%. Furthermore, student LMs fine-tuned with rationales extracted via DToT outperform baselines on all datasets with up to 16.9% accuracy improvement, while being more than 60x smaller than conventional LLMs. Finally, we observe that student LMs fine-tuned with rationales exhibit better cross-dataset transferability.

We consider the problem of predicting cellular network performance (signal maps) from measurements collected by several mobile devices. We formulate the problem within the online federated learning framework: (i) federated learning (FL) enables users to collaboratively train a model, while keeping their training data on their devices; (ii) measurements are collected as users move around over time and are used for local training in an online fashion. We consider an honest-but-curious server, who observes the updates from target users participating in FL and infers their location using a deep leakage from gradients (DLG) type of attack, originally developed to reconstruct training data of DNN image classifiers. We make the key observation that a DLG attack, applied to our setting, infers the average location of a batch of local data, and can thus be used to reconstruct the target users' trajectory at a coarse granularity. We build on this observation to protect location privacy, in our setting, by revisiting and designing mechanisms within the federated learning framework including: tuning the FL parameters for averaging, curating local batches so as to mislead the DLG attacker, and aggregating across multiple users with different trajectories. We evaluate the performance of our algorithms through both analysis and simulation based on real-world mobile datasets, and we show that they achieve a good privacy-utility tradeoff.

Federated learning (FL) has attracted growing interest for enabling privacy-preserving machine learning on data stored at multiple users while avoiding moving the data off-device. However, while data never leaves users’ devices, privacy still cannot be guaranteed since significant computations on users’ training data are shared in the form of trained local models. These local models have recently been shown to pose a substantial privacy threat through different privacy attacks such as model inversion attacks. As a remedy, Secure Aggregation (SA) has been developed as a framework to preserve privacy in FL, by guaranteeing the server can only learn the global aggregated model update but not the individual model updates.While SA ensures no additional information is leaked about the individual model update beyond the aggregated model update, there are no formal guarantees on how much privacy FL with SA can actually offer; as information about the individual dataset can still potentially leak through the aggregated model computed at the server. In this work, we perform a first analysis of the formal privacy guarantees for FL with SA. Specifically, we use Mutual Information (MI) as a quantification metric and derive upper bounds on how much information about each user's dataset can leak through the aggregated model update. When using the FedSGD aggregation algorithm, our theoretical bounds show that the amount of privacy leakage reduces linearly with the number of users participating in FL with SA. To validate our theoretical bounds, we use an MI Neural Estimator to empirically evaluate the privacy leakage under different FL setups on both the MNIST and CIFAR10 datasets. Our experiments verify our theoretical bounds for FedSGD, which show a reduction in privacy leakage as the number of users and local batch size grow, and an increase in privacy leakage as the number of training rounds increases. We also observe similar dependencies for the FedAvg and FedProx protocol.

Online content platforms optimize engagement by providing personalized recommendations to their users. These recommendation systems track and profile users to predict relevant content a user is likely interested in. While the personalized recommendations provide utility to users, the tracking and profiling that enables them poses a privacy issue because the platform might infer potentially sensitive user interests. There is increasing interest in building privacy-enhancing obfuscation approaches that do not rely on cooperation from online content platforms. However, existing obfuscation approaches primarily focus on enhancing privacy but at the same time they degrade the utility because obfuscation introduces unrelated recommendations. We design and implement DeHarpo, an obfuscation approach for YouTube's recommendation system that not only obfuscates a user's video watch history to protect privacy but then also denoises the video recommendations by YouTube to preserve their utility. In contrast to prior obfuscation approaches, DeHarpo adds a denoiser that makes use of a ``secret'' input (i.e., a user's actual watch history) as well as information that is also available to the adversarial recommendation system (i.e., obfuscated watch history and corresponding ``nois`` recommendations). Our large-scale evaluation of DeHarpo shows that it outperforms the state-of-the-art by a factor of 2x in terms of preserving utility for the same level of privacy, while maintaining stealthiness and robustness to de-obfuscation.

Emerging technologies and applications have generated large amounts of data at the network edge. Due to bandwidth, storage, and privacy concerns, it is often impractical to move the collected data to the cloud. With the rapid development of edge computing and distributed machine learning (ML), edge-based ML called federated learning has emerged to overcome the shortcomings of cloud-based ML. Existing works mainly focus on designing efficient learning algorithms, few works focus on designing the incentive mechanisms with heterogeneous edge nodes (EN) and uncertainty of network bandwidth. The incentive mechanisms affect various tradeoffs: (i) between computation and communication latency, and thus (ii) between the edge learning time and payment consumption. We fill this gap by designing an incentive mechanism that captures the tradeoff between latency and payment. Due to the network dynamics and privacy protection, we propose a deep reinforcement learning-based (DRL-based) solution that can automatically learn the best pricing strategy. To the best of our knowledge, this is the first work that applies the advances of DRL to design the incentive mechanism for edge learning. We evaluate the performance of the incentive mechanism using trace-driven experiments. The results demonstrate the superiority of our proposed approach as compared with the baselines.

Edge computing is a new paradigm to provide strong computing capability at the edge of pervasive radio access networks close to users. A critical research challenge of edge computing is to design an efficient offloading strategy to decide which tasks can be offloaded to edge servers with limited resources. Although many research efforts attempt to address this challenge, they need centralized control, which is not practical because users are rational individuals with interests to maximize their benefits. In this article, we study to design a decentralized algorithm for computation offloading, so that users can independently choose their offloading decisions. Game theory has been applied in the algorithm design. Different from existing work, we address the challenge that users may refuse to expose their information about network bandwidth and preference. Therefore, it requires that our solution should make the offloading decision without such knowledge. We formulate the problem as a partially observable Markov decision process (POMDP), which is solved by a policy gradient deep reinforcement learning (DRL) based approach. Extensive simulation results show that our proposal significantly outperforms existing solutions.

The explosive increase of mobile devices with built-in sensors such as GPS, accelerometer, gyroscope and camera has made the design of mobile crowdsensing (MCS) applications possible, which create a new interface between humans and their surroundings. Until now, various MCS applications have been designed, where the task initiators (TIs) recruit mobile users (MUs) to complete the required sensing tasks. In this paper, deep reinforcement learning (DRL) based techniques are investigated to address the problem of assigning satisfactory but profitable amount of incentives to multiple TIs and MUs as a MCS game. Specifically, we first formulate the problem as a multi-leader and multi-follower Stackelberg game, where TIs are the leaders and MUs are the followers. Then, the existence of the Stackelberg Equilibrium (SE) is proved. Considering the challenge to compute the SE, a DRL based Dynamic Incentive Mechanism (DDIM) is proposed. It enables the TIs to learn the optimal pricing strategies directly from game experiences without knowing the private information of MUs. Finally, numerical experiments are provided to illustrate the effectiveness of the proposed incentive mechanism compared with both state-of-the-art and baseline approaches.