MoE-TransMov: A Transformer-based Model for Next POI Prediction in Familiar & Unfamiliar Movements

Accurate prediction of the next point of interest (POI) within human mobility trajectories is essential for location-based services, as it enables more timely and personalized recommendations. In particular, with the rise of these approaches, studies have shown that users exhibit different POI choices in their familiar and unfamiliar areas, highlighting the importance of incorporating user familiarity into predictive models. However, existing methods often fail to distinguish between the movements of users in familiar and unfamiliar regions. To address this, we propose MoE-TransMov, a Transformer-based model with a Transformer model with a Mixture-of-Experts (MoE) architecture designed to use one framework to capture distinct mobility patterns across different moving contexts without requiring separate training for certain data. Using user-check-in data, we classify movements into familiar and unfamiliar categories and develop a specialized expert network to improve prediction accuracy. Our approach integrates self-attention mechanisms and adaptive gating networks to dynamically select the most relevant expert models for different mobility contexts. Experiments on two real-world datasets, including the widely used but small open-source Foursquare NYC dataset and the large-scale Kyoto dataset collected with LY Corporation (Yahoo Japan Corporation), show that MoE-TransMov outperforms state-of-the-art baselines with notable improvements in Top-1, Top-5, Top-10 accuracy, and mean reciprocal rank (MRR). Given the results, we find that by using this approach, we can efficiently improve mobility predictions under different moving contexts, thereby enhancing the personalization of recommendation systems and advancing various urban applications.

The Role of Islamic Ethics in Preventing the Abuse of Artificial Intelligence (AI) Based Deepfakes

The significant development of deepfake technology powered by artificial intelligence (AI) has sparked worldwide concerns about the alteration of false information, the usurpation of online identities, and the decline of public confidence in the authenticity of online content. These incidents not only raise technical issues but also carry complex moral implications, rendering conventional, technologically driven, and reactive management methods inadequate to address the underlying causes of the problem, including intent, morality, and potential intangible social impacts. Based on these issues, this study aims to formulate a comprehensive Islamic ethical framework that can serve as a more comprehensive preventative tool to mitigate the risks of misuse of deepfakes. The study employed a Systematic Literature Review (SLR) guided by PRISMA, selecting ten primary sources published between 2018 and 2025 to identify ethical deficiencies, regulatory needs, and appropriate normative solutions. The analysis shows that the integration of the principles of (Maqasid al-Shariah) particularly (hifz al-ird) protecting honor and (hifz al-nafs) protecting the self, provides a strong normative basis for regulating the responsible use of technology. This study yields three strategic recommendations: regulatory changes that recognize the intangible and psychological harm caused by reputational damage; improved technology management through moral scrutiny that upholds the values of justice (adl), trust, and openness; and increased public digital literacy based on the principle of (tabayyun) examination and caution. Overall, this study concludes that the application of Islamic ethics offers a shift in thinking from punitive mechanisms to preventative approaches that focus on protecting human dignity, preventing harm, and strengthening the common good in the digital age.

A Systematic Reproducibility Study of BSARec for Sequential Recommendation

In sequential recommendation (SR), the self-attention mechanism of Transformer-based models acts as a low-pass filter, limiting their ability to capture high-frequency signals that reflect short-term user interests. To overcome this, BSARec augments the Transformer encoder with a frequency layer that rescales high-frequency components using the Fourier transform. However, the overall effectiveness of BSARec and the roles of its individual components have yet to be systematically validated. We reproduce BSARec and show that it outperforms other SR methods on some datasets. To empirically assess whether BSARec improves performance on high-frequency signals, we propose a metric to quantify user history frequency and evaluate SR methods across different user groups. We compare digital signal processing (DSP) techniques and find that the discrete wavelet transform (DWT) offer only slight improvements over Fourier transforms, and DSP methods provide no clear advantage over simple residual connections. Finally, we explore padding strategies and find that non-constant padding significantly improves recommendation performance, whereas constant padding hinders the frequency rescaler's ability to capture high-frequency signals.

Exploiting ID-Text Complementarity via Ensembling for Sequential Recommendation

Modern Sequential Recommendation (SR) models commonly utilize modality features to represent items, motivated in large part by recent advancements in language and vision modeling. To do so, several works completely replace ID embeddings with modality embeddings, claiming that modality embeddings render ID embeddings unnecessary because they can match or even exceed ID embedding performance. On the other hand, many works jointly utilize ID and modality features, but posit that complex fusion strategies, such as multi-stage training and/or intricate alignment architectures, are necessary for this joint utilization. However, underlying both these lines of work is a lack of understanding of the complementarity of ID and modality features. In this work, we address this gap by studying the complementarity of ID- and text-based SR models. We show that these models do learn complementary signals, meaning that either should provide performance gain when used properly alongside the other. Motivated by this, we propose a new SR method that preserves ID-text complementarity through independent model training, then harnesses it through a simple ensembling strategy. Despite this method's simplicity, we show it outperforms several competitive SR baselines, implying that both ID and text features are necessary to achieve state-of-the-art SR performance but complex fusion architectures are not.

Holistic Evaluation of State-of-the-Art LLMs for Code Generation

This study presents a comprehensive empirical evaluation of six state-of-the-art large language models (LLMs) for code generation, including both general-purpose and code-specialized models. Using a dataset of 944 real-world LeetCode problems across five programming languages, we assess model performance using rigorous metrics: compile-time errors, runtime errors, functional failures, and algorithmic suboptimalities. The results reveal significant performance variations, with DeepSeek-R1 and GPT-4.1 consistently outperform others in terms of correctness, efficiency, and robustness. Through detailed case studies, we identify common failure scenarios such as syntax errors, logical flaws, and suboptimal algorithms, highlighting the critical role of prompt engineering and human oversight in improving results. Based on these findings, we provide actionable recommendations for developers and practitioners, emphasizing that successful LLM deployment depends on careful model selection, effective prompt design, and context-aware usage to ensure reliable code generation in real-world software development tasks.

Probabilistic Digital Twins of Users: Latent Representation Learning with Statistically Validated Semantics

Understanding user identity and behavior is central to applications such as personalization, recommendation, and decision support. Most existing approaches rely on deterministic embeddings or black-box predictive models, offering limited uncertainty quantification and little insight into what latent representations encode. We propose a probabilistic digital twin framework in which each user is modeled as a latent stochastic state that generates observed behavioral data. The digital twin is learned via amortized variational inference, enabling scalable posterior estimation while retaining a fully probabilistic interpretation. We instantiate this framework using a variational autoencoder (VAE) applied to a user-response dataset designed to capture stable aspects of user identity. Beyond standard reconstruction-based evaluation, we introduce a statistically grounded interpretation pipeline that links latent dimensions to observable behavioral patterns. By analyzing users at the extremes of each latent dimension and validating differences using nonparametric hypothesis tests and effect sizes, we demonstrate that specific dimensions correspond to interpretable traits such as opinion strength and decisiveness. Empirically, we find that user structure is predominantly continuous rather than discretely clustered, with weak but meaningful structure emerging along a small number of dominant latent axes. These results suggest that probabilistic digital twins can provide interpretable, uncertainty-aware representations that go beyond deterministic user embeddings.

The Mental World of Large Language Models in Recommendation: A Benchmark on Association, Personalization, and Knowledgeability

Large language models (LLMs) have shown potential in recommendation systems (RecSys) by using them as either knowledge enhancer or zero-shot ranker. A key challenge lies in the large semantic gap between LLMs and RecSys where the former internalizes language world knowledge while the latter captures personalized world of behaviors. Unfortunately, the research community lacks a comprehensive benchmark that evaluates the LLMs over their limitations and boundaries in RecSys so that we can draw a confident conclusion. To investigate this, we propose a benchmark named LRWorld containing over 38K high-quality samples and 23M tokens carefully compiled and generated from widely used public recommendation datasets. LRWorld categorizes the mental world of LLMs in RecSys as three main scales (association, personalization, and knowledgeability) spanned by ten factors with 31 measures (tasks). Based on LRWorld, comprehensive experiments on dozens of LLMs show that they are still not well capturing the deep neural personalized embeddings but can achieve good results on shallow memory-based item-item similarity. They are also good at perceiving item entity relations, entity hierarchical taxonomies, and item-item association rules when inferring user interests. Furthermore, LLMs show a promising ability in multimodal knowledge reasoning (movie poster and product image) and robustness to noisy profiles. None of them show consistently good performance over the ten factors. Model sizes, position bias, and more are ablated.

Warmer for Less: A Cost-Efficient Strategy for Cold-Start Recommendations at Pinterest

Pinterest is a leading visual discovery platform where recommender systems (RecSys) are key to delivering relevant, engaging, and fresh content to our users. In this paper, we study the problem of improving RecSys model predictions for cold-start (CS) items, which appear infrequently in the training data. Although this problem is well-studied in academia, few studies have addressed its root causes effectively at the scale of a platform like Pinterest. By investigating live traffic data, we identified several challenges of the CS problem and developed a corresponding solution for each: First, industrial-scale RecSys models must operate under tight computational constraints. Since CS items are a minority, any related improvements must be highly cost-efficient. To address this, our solutions were designed to be lightweight, collectively increasing the total parameters by only 5%. Second, CS items are represented only by non-historical (e.g., content or attribute) features, which models often treat as less important. To elevate their significance, we introduce a residual connection for the non-historical features. Third, CS items tend to receive lower prediction scores compared to non-CS items, reducing their likelihood of being surfaced. We mitigate this by incorporating a score regularization term into the model. Fourth, the labels associated with CS items are sparse, making it difficult for the model to learn from them. We apply the manifold mixup technique to address this data sparsity. Implemented together, our methods increased fresh content engagement at Pinterest by 10% without negatively impacting overall engagement and cost, and have been deployed to serve over 570 million users on Pinterest.

Quantifying Laziness, Decoding Suboptimality, and Context Degradation in Large Language Models

Large Language Models (LLMs) often exhibit behavioral artifacts such as laziness (premature truncation of responses or partial compliance with multi-part requests), decoding suboptimality (failure to select higher-quality sequences due to myopic decoding), and context degradation (forgetting or ignoring core instructions over long conversations). We conducted three controlled experiments (A, B, and C) to quantify these phenomena across several advanced LLMs (OpenAI GPT-4 variant, DeepSeek). Our results indicate widespread laziness in satisfying complex multi-part instructions: models frequently omitted required sections or failed to meet length requirements despite explicit prompting. However, we found limited evidence of decoding suboptimality in a simple reasoning task (the models' greedy answers appeared to align with their highest-confidence solution), and we observed surprising robustness against context degradation in a 200-turn chaotic conversation test - the models maintained key facts and instructions far better than expected. These findings suggest that while compliance with detailed instructions remains an open challenge, modern LLMs may internally mitigate some hypothesized failure modes (such as context forgetting) in straightforward retrieval scenarios. We discuss implications for reliability, relate our findings to prior work on instruction-following and long-context processing, and recommend strategies (such as self-refinement and dynamic prompting) to reduce laziness and bolster multi-instruction compliance.

xGR: Efficient Generative Recommendation Serving at Scale

Recommendation system delivers substantial economic benefits by providing personalized predictions. Generative recommendation (GR) integrates LLMs to enhance the understanding of long user-item sequences. Despite employing attention-based architectures, GR's workload differs markedly from that of LLM serving. GR typically processes long prompt while producing short, fixed-length outputs, yet the computational cost of each decode phase is especially high due to the large beam width. In addition, since the beam search involves a vast item space, the sorting overhead becomes particularly time-consuming. We propose xGR, a GR-oriented serving system that meets strict low-latency requirements under highconcurrency scenarios. First, xGR unifies the processing of prefill and decode phases through staged computation and separated KV cache. Second, xGR enables early sorting termination and mask-based item filtering with data structure reuse. Third, xGR reconstructs the overall pipeline to exploit multilevel overlap and multi-stream parallelism. Our experiments with real-world recommendation service datasets demonstrate that xGR achieves at least 3.49x throughput compared to the state-of-the-art baseline under strict latency constraints.