Ming Li

Yuhang Zhou, Jing Zhu, Shengyi Qian et al.

Large Language Models (LLMs) are increasingly aligned with human preferences through Reinforcement Learning from Human Feedback (RLHF). Among RLHF methods, Group Relative Policy Optimization (GRPO) has gained attention for its simplicity and strong performance, notably eliminating the need for a learned value function. However, GRPO implicitly assumes a balanced domain distribution and uniform semantic alignment across groups, assumptions that rarely hold in real-world datasets. When applied to multi-domain, imbalanced data, GRPO disproportionately optimizes for dominant domains, neglecting underrepresented ones and resulting in poor generalization and fairness. We propose Domain-Informed Self-Consistency Policy Optimization (DISCO), a principled extension to GRPO that addresses inter-group imbalance with two key innovations. Domain-aware reward scaling counteracts frequency bias by reweighting optimization based on domain prevalence. Difficulty-aware reward scaling leverages prompt-level self-consistency to identify and prioritize uncertain prompts that offer greater learning value. Together, these strategies promote more equitable and effective policy learning across domains. Extensive experiments across multiple LLMs and skewed training distributions show that DISCO improves generalization, outperforms existing GRPO variants by 5% on Qwen3 models, and sets new state-of-the-art results on multi-domain alignment benchmarks. Our code and data are available at https://github.com/Tonyzhou98/disco_grpo.

Haichen Wang, Liu Yang, Xinyuan Zhang et al.

Passenger demand forecasting helps optimize vehicle scheduling, thereby improving urban efficiency. Recently, attention-based methods have been used to adequately capture the dynamic nature of spatio-temporal data. However, existing methods that rely on heuristic masking strategies cannot fully adapt to the complex spatio-temporal correlations, hindering the model from focusing on the right context. These works also overlook the high-level correlations that exist in the real world. Effectively integrating these high-level correlations with the original correlations is crucial. To fill this gap, we propose the Aggregation Differential Transformer (ADFormer), which offers new insights to demand forecasting promotion. Specifically, we utilize Differential Attention to capture the original spatial correlations and achieve attention denoising. Meanwhile, we design distinct aggregation strategies based on the nature of space and time. Then, the original correlations are unified with the high-level correlations, enabling the model to capture holistic spatio-temporal relations. Experiments conducted on taxi and bike datasets confirm the effectiveness and efficiency of our model, demonstrating its practical value. The code is available at https://github.com/decisionintelligence/ADFormer.

Xiyao Wang, Zhengyuan Yang, Chao Feng et al. · microsoft-research

Reinforcement learning (RL) has shown great effectiveness for fine-tuning large language models (LLMs) using tasks that are challenging yet easily verifiable, such as math reasoning or code generation. However, extending this success to visual perception in vision-language models (VLMs) has been impeded by the scarcity of vision-centric tasks that are simultaneously challenging and unambiguously verifiable. To this end, we introduce ViCrit (Visual Caption Hallucination Critic), an RL proxy task that trains VLMs to localize a subtle, synthetic visual hallucination injected into paragraphs of human-written image captions. Starting from a 200-word captions, we inject a single, subtle visual description error-altering a few words on objects, attributes, counts, or spatial relations-and task the model to pinpoint the corrupted span given the image and the modified caption. This formulation preserves the full perceptual difficulty while providing a binary, exact-match reward that is easy to compute and unambiguous. Models trained with the ViCrit Task exhibit substantial gains across a variety of VL benchmarks. Crucially, the improvements transfer beyond natural-image training data to abstract image reasoning and visual math, showing promises of learning to perceive rather than barely memorizing seen objects. To facilitate evaluation, we further introduce ViCrit-Bench, a category-balanced diagnostic benchmark that systematically probes perception errors across diverse image domains and error types. Together, our results demonstrate that fine-grained hallucination criticism is an effective and generalizable objective for enhancing visual perception in VLMs.

Simin Wang, Liguo Huang, Jidong Ge et al.

Since 2009, the deep learning revolution, which was triggered by the introduction of ImageNet, has stimulated the synergy between Machine Learning (ML)/Deep Learning (DL) and Software Engineering (SE). Meanwhile, critical reviews have emerged that suggest that ML/DL should be used cautiously. To improve the quality (especially the applicability and generalizability) of ML/DL-related SE studies, and to stimulate and enhance future collaborations between SE/AI researchers and industry practitioners, we conducted a 10-year Systematic Literature Review (SLR) on 906 ML/DL-related SE papers published between 2009 and 2018. Our trend analysis demonstrated the mutual impacts that ML/DL and SE have had on each other. At the same time, however, we also observed a paucity of replicable and reproducible ML/DL-related SE studies and identified five factors that influence their replicability and reproducibility. To improve the applicability and generalizability of research results, we analyzed what ingredients in a study would facilitate an understanding of why a ML/DL technique was selected for a specific SE problem. In addition, we identified the unique trends of impacts of DL models on SE tasks, as well as five unique challenges that needed to be met in order to better leverage DL to improve the productivity of SE tasks. Finally, we outlined a road-map that we believe can facilitate the transfer of ML/DL-based SE research results into real-world industry practices.

Ya-Lin Zhang, Jun Zhou, Wenhao Zheng et al.

Internet companies are facing the need for handling large-scale machine learning applications on a daily basis and distributed implementation of machine learning algorithms which can handle extra-large scale tasks with great performance is widely needed. Deep forest is a recently proposed deep learning framework which uses tree ensembles as its building blocks and it has achieved highly competitive results on various domains of tasks. However, it has not been tested on extremely large scale tasks. In this work, based on our parameter server system, we developed the distributed version of deep forest. To meet the need for real-world tasks, many improvements are introduced to the original deep forest model, including MART (Multiple Additive Regression Tree) as base learners for efficiency and effectiveness consideration, the cost-based method for handling prevalent class-imbalanced data, MART based feature selection for high dimension data and different evaluation metrics for automatically determining of the cascade level. We tested the deep forest model on an extra-large scale task, i.e., automatic detection of cash-out fraud, with more than 100 millions of training samples. Experimental results showed that the deep forest model has the best performance according to the evaluation metrics from different perspectives even with very little effort for parameter tuning. This model can block fraud transactions in a large amount of money each day. Even compared with the best-deployed model, the deep forest model can additionally bring into a significant decrease in economic loss each day.