Dongze Wu

Maolin Wang, Dongze Wu, Jianing Zhou et al.

Large language models (LLMs) have become an important semantic infrastructure for modern recommender systems. A prevailing paradigm integrates LLM-derived semantic embeddings with collaborative representations via representation alignment, implicitly assuming that the two views encode a shared latent entity and that stronger alignment yields better results. We formalize this assumption as the global low-complexity alignment hypothesis and argue that it is stronger than necessary and often structurally mismatched with real-world recommendation settings. We propose a complementary perspective in which semantic and collaborative representations are treated as partially shared yet fundamentally heterogeneous views, each containing both shared and view-specific factors. Under this shared-plus-private latent structure, enforcing global geometric alignment may distort local structure, suppress view-specific signals, and reduce informational diversity. To support this perspective, we develop complementarity-aware diagnostics that quantify overlap, unique-hit contribution, and theoretical fusion upper bounds. Empirical analyses on sparse recommendation benchmarks reveal low item-level agreement between semantic and collaborative views and substantial oracle fusion gains, indicating strong complementarity. Furthermore, controlled alignment probes show that low-capacity mappings capture only shared components and fail to recover full collaborative geometry, especially under distribution shift. These findings suggest that alignment should not be treated as the default integration principle. We advocate a shift from alignment-centric modeling to complementarity fusion-centric, complementarity-aware design, where shared factors are selectively integrated while private signals are preserved. This reframing provides a principled foundation for the next generation of LLM-enhanced recommender systems.

Dongze Wu, Yao Xie

Sampling from high-dimensional, multi-modal distributions remains a fundamental challenge across domains such as statistical Bayesian inference and physics-based machine learning. In this paper, we propose Annealing Flow (AF), a method built on Continuous Normalizing Flow (CNF) for sampling from high-dimensional and multi-modal distributions. AF is trained with a dynamic Optimal Transport (OT) objective incorporating Wasserstein regularization, and guided by annealing procedures, facilitating effective exploration of modes in high-dimensional spaces. Compared to recent NF methods, AF greatly improves training efficiency and stability, with minimal reliance on MC assistance. We demonstrate the superior performance of AF compared to state-of-the-art methods through experiments on various challenging distributions and real-world datasets, particularly in high-dimensional and multi-modal settings. We also highlight AF potential for sampling the least favorable distributions.

Dongze Wu, David I. Inouye, Yao Xie

Predicting potential and counterfactual outcomes from observational data is central to clinical decision-making, where physicians must weigh treatments for an individual patient rather than relying solely on average effects at the population level. We propose PO-Flow, a continuous normalizing flow (CNF) framework for causal inference that jointly models potential outcomes and counterfactuals. Trained via flow matching, PO-Flow provides a unified approach to average treatment effect estimation, individualized potential outcome prediction, and counterfactual prediction. Besides, PO-Flow directly learns the densities of potential outcomes, enabling likelihood-based evaluation of predictions. Furthermore, PO-Flow explores counterfactual outcome generation conditioned on the observed factual in general observational datasets, with a supporting recovery result under certain assumptions. PO-Flow outperforms modern baselines across diverse datasets and causal tasks in the potential outcomes framework.

Dongze Wu, Hanyang Jiang, Yao Xie

Bias in data collection, arising from both under-reporting and over-reporting, poses significant challenges in critical applications such as healthcare and public safety. In this work, we introduce Graph-based Over- and Under-reporting Debiasing (GROUD), a novel graph-based optimization framework that debiases reported data by jointly estimating the true incident counts and the associated reporting bias probabilities. By modeling the bias as a smooth signal over a graph constructed from geophysical or feature-based similarities, our convex formulation not only ensures a unique solution but also comes with theoretical recovery guarantees under certain assumptions. We validate GROUD on both challenging simulated experiments and real-world datasets -- including Atlanta emergency calls and COVID-19 vaccine adverse event reports -- demonstrating its robustness and superior performance in accurately recovering debiased counts. This approach paves the way for more reliable downstream decision-making in systems affected by reporting irregularities.