Bid shading plays a crucial role in Real-Time Bidding (RTB) by adaptively adjusting the bid to avoid advertisers overspending. Existing mainstream two-stage methods, which first model bid landscapes and then optimize surplus using operations research techniques, are constrained by unimodal assumptions that fail to adapt for non-convex surplus curves and are vulnerable to cascading errors in sequential workflows. Additionally, existing discretization models of continuous values ignore the dependence between discrete intervals, reducing the model's error correction ability, while sample selection bias in bidding scenarios presents further challenges for prediction. To address these issues, this paper introduces Generative Bid Shading (GBS), which comprises two primary components: 1) an end-to-end generative model that utilizes an autoregressive approach to generate shading ratios by stepwise residuals, capturing complex value dependencies without relying on predefined priors; and 2) a reward preference alignment system, which incorporates a channel-aware hierarchical dynamic network (CHNet) as the reward model to extract fine-grained features, along with modules for surplus optimization and exploration utility reward alignment, ultimately optimizing both short-term and long-term surplus using group relative policy optimization (GRPO). Extensive experiments on both offline and online A/B tests validate GBS's effectiveness. Moreover, GBS has been deployed on the Meituan DSP platform, serving billions of bid requests daily.

Reranking plays a crucial role in modern recommender systems by capturing the mutual influences within the list. Due to the inherent challenges of combinatorial search spaces, most methods adopt a two-stage search paradigm: a simple General Search Unit (GSU) efficiently reduces the candidate space, and an Exact Search Unit (ESU) effectively selects the optimal sequence. These methods essentially involve making trade-offs between effectiveness and efficiency, while suffering from a severe inconsistency problem, that is, the GSU often misses high-value lists from ESU. To address this problem, we propose YOLOR, a one-stage reranking method that removes the GSU while retaining only the ESU. Specifically, YOLOR includes: (1) a Tree-based Context Extraction Module (TCEM) that hierarchically aggregates multi-scale contextual features to achieve ''list-level effectiveness'', and (2) a Context Cache Module (CCM) that enables efficient feature reuse across candidate permutations to achieve ''permutation-level efficiency''. Extensive experiments across public and industry datasets validate YOLOR's performance and we have successfully deployed YOLOR on the Meituan food delivery platform.

The rapid proliferation of fake news on social media threatens social stability, creating an urgent demand for more effective detection methods. While many promising approaches have emerged, most rely on content analysis with limited semantic depth, leading to suboptimal comprehension of news content. To address this limitation, capturing broader-range semantics is essential yet challenging, as it introduces two primary types of noise: fully connecting sentences in news graphs often adds unnecessary structural noise, while highly similar but authenticity-irrelevant sentences introduce feature noise, complicating the detection process. To tackle these issues, we propose BREAK, a broad-range semantics model for fake news detection that leverages a fully connected graph to capture comprehensive semantics while employing dual denoising modules to minimize both structural and feature noise. The semantic structure denoising module balances the graph's connectivity by iteratively refining it between two bounds: a sequence-based structure as a lower bound and a fully connected graph as the upper bound. This refinement uncovers label-relevant semantic interrelations structures. Meanwhile, the semantic feature denoising module reduces noise from similar semantics by diversifying representations, aligning distinct outputs from the denoised graph and sequence encoders using KL-divergence to achieve feature diversification in high-dimensional space. The two modules are jointly optimized in a bi-level framework, enhancing the integration of denoised semantics into a comprehensive representation for detection. Extensive experiments across four datasets prove that BREAK significantly outperforms existing fake news detection methods.

Modern social media’s development has dramatically changed how people obtain information. However, the wide dissemination of various false information has severe detrimental effects. Accordingly, many deep learning-based methods have been proposed to detect false information and achieve promising results. However, these methods are unsuitable for new events due to the extremely limited labeled data and their discrepant data distribution to existing events. Domain adaptation methods have been proposed to mitigate these problems. However, their performance is suboptimal because they are not sensitive to new events due to they aim to align the domain information between existing events, and they hardly capture the fine-grained difference between real and fake claims by only using semantic information. Therefore, we propose a novel Emotion-aware Meta Learning (EML) approach for cross-event false information early detection, which deeply integrates emotions in meta learning to find event-sensitive initialization parameters that quickly adapt to new events. EML is non-trivial and faces three challenges: (1) How to effectively model semantic and emotional features to capture fine-grained differences? (2) How to reduce the impact of noise in meta learning based on semantic and emotional features? (3) How to detect the false information in a zero-shot detection scenario, i.e., no labeled data for new events? To tackle these challenges, firstly, we construct the emotion-aware meta tasks by selecting claims with similar and opposite emotions to the target claim other than usually used random sampling. Secondly, we propose a task weighting method and event-adaptation meta tasks to further improve the model’s robustness and generalization ability for detecting new events. Finally, we propose a weak label annotation method to extend EML to zero-shot detection according to the calculated labels’ confidence. Extensive experiments on real-world datasets show that the EML achieves superior performances on false information detection for new events.