Chang Tian

Chang Tian, Wenpeng Yin, Marie-Francine Moens

A dialogue policy module is an essential part of task-completion dialogue systems. Recently, increasing interest has focused on reinforcement learning (RL)-based dialogue policy. Its favorable performance and wise action decisions rely on an accurate estimation of action values. The overestimation problem is a widely known issue of RL since its estimate of the maximum action value is larger than the ground truth, which results in an unstable learning process and suboptimal policy. This problem is detrimental to RL-based dialogue policy learning. To mitigate this problem, this paper proposes a dynamic partial average estimator (DPAV) of the ground truth maximum action value. DPAV calculates the partial average between the predicted maximum action value and minimum action value, where the weights are dynamically adaptive and problem-dependent. We incorporate DPAV into a deep Q-network as the dialogue policy and show that our method can achieve better or comparable results compared to top baselines on three dialogue datasets of different domains with a lower computational load. In addition, we also theoretically prove the convergence and derive the upper and lower bounds of the bias compared with those of other methods.

Ke Ma, Yiliang Sang, Yang Ming et al.

The Type-II codebook in Release 17 (R17) exploits the angular-delay-domain partial reciprocity between uplink and downlink channels to select part of angular-delay-domain ports for measuring and feeding back the downlink channel state information (CSI), where the performance of existing deep learning enhanced CSI feedback methods is limited due to the deficiency of sparse structures. To address this issue, we propose two new perspectives of adopting deep learning to improve the R17 Type-II codebook. Firstly, considering the low signal-to-noise ratio of uplink channels, deep learning is utilized to accurately select the dominant angular-delay-domain ports, where the focal loss is harnessed to solve the class imbalance problem. Secondly, we propose to adopt deep learning to reconstruct the downlink CSI based on the feedback of the R17 Type-II codebook at the base station, where the information of sparse structures can be effectively leveraged. Besides, a weighted shortcut module is designed to facilitate the accurate reconstruction. Simulation results demonstrate that our proposed methods could improve the sum rate performance compared with its traditional R17 Type-II codebook and deep learning benchmarks.

Fan Zhou, Chang Tian, Tim Van de Cruys

Generating stylistic text with specific attributes is a key problem in controllable text generation. Recently, diffusion models have emerged as a powerful paradigm for both visual and textual generation. Existing approaches can be broadly categorized into classifier-free guidance (CFG) and classifier guidance (CG) methods. While CFG effectively preserves semantic content, it often fails to provide effective attribute control. In contrast, CG modifies the denoising trajectory using classifier gradients, enabling better attribute alignment but incurring high computational costs during sampling and suffering from classifier generalization issues. In this work, we propose RegDiff, a regularized diffusion framework that leverages attribute features without requiring a pretrained classifier during sampling, thereby achieving controllable generation with reduced computational costs. Specifically, RegDiff employs a VAE-based encoder--decoder architecture to ensure reconstruction fidelity and a latent diffusion model trained with attribute supervision to enable controllable text generation. Attribute information is injected only during training. Experiments on five datasets spanning multiple stylistic attributes demonstrate that RegDiff outperforms strong baselines in generating stylistic texts. These results validate the effectiveness of RegDiff as an efficient solution for attribute-controllable text diffusion. Our code, datasets, and resources will be released upon publication at https://github.com/xxxx.

Chang Tian, Matthew B. Blaschko, Wenpeng Yin et al.

Fine-grained category discovery using only coarse-grained supervision is a cost-effective yet challenging task. Previous training methods focus on aligning query samples with positive samples and distancing them from negatives. They often neglect intra-category and inter-category semantic similarities of fine-grained categories when navigating sample distributions in the embedding space. Furthermore, some evaluation techniques that rely on pre-collected test samples are inadequate for real-time applications. To address these shortcomings, we introduce a method that successfully detects fine-grained clusters of semantically similar texts guided by a novel objective function. The method uses semantic similarities in a logarithmic space to guide sample distributions in the Euclidean space and to form distinct clusters that represent fine-grained categories. We also propose a centroid inference mechanism to support real-time applications. The efficacy of the method is both theoretically justified and empirically confirmed on three benchmark tasks. The proposed objective function is integrated in multiple contrastive learning based neural models. Its results surpass existing state-of-the-art approaches in terms of Accuracy, Adjusted Rand Index and Normalized Mutual Information of the detected fine-grained categories. Code and data will be available at Code and data are publicly available at https://github.com/changtianluckyforever/F-grained-STAR.

Dan Li, Shuai Wang, Jie Zou et al.

In this work we propose a new task: artistic visualization of classical Chinese poems, where the goal is to generatepaintings of a certain artistic style for classical Chinese poems. For this purpose, we construct a new dataset called Paint4Poem. Thefirst part of Paint4Poem consists of 301 high-quality poem-painting pairs collected manually from an influential modern Chinese artistFeng Zikai. As its small scale poses challenges for effectively training poem-to-painting generation models, we introduce the secondpart of Paint4Poem, which consists of 3,648 caption-painting pairs collected manually from Feng Zikai's paintings and 89,204 poem-painting pairs collected automatically from the web. We expect the former to help learning the artist painting style as it containshis most paintings, and the latter to help learning the semantic relevance between poems and paintings. Further, we analyze Paint4Poem regarding poem diversity, painting style, and the semantic relevance between poems and paintings. We create abenchmark for Paint4Poem: we train two representative text-to-image generation models: AttnGAN and MirrorGAN, and evaluate theirperformance regarding painting pictorial quality, painting stylistic relevance, and semantic relevance between poems and paintings.The results indicate that the models are able to generate paintings that have good pictorial quality and mimic Feng Zikai's style, but thereflection of poem semantics is limited. The dataset also poses many interesting research directions on this task, including transferlearning, few-shot learning, text-to-image generation for low-resource data etc. The dataset is publicly available.(https://github.com/paint4poem/paint4poem)

Chang Tian, Matthew B. Blaschko, Mingzhe Xing et al.

Reinforcement learning (RL) has become a key technique for enhancing the reasoning abilities of large language models (LLMs), with policy-gradient algorithms dominating the post-training stage because of their efficiency and effectiveness. However, most existing benchmarks evaluate large-language-model reasoning under idealized settings, overlooking performance in realistic, non-ideal scenarios. We identify three representative non-ideal scenarios with practical relevance: summary inference, fine-grained noise suppression, and contextual filtering. We introduce a new research direction guided by brain-science findings that human reasoning remains reliable under imperfect inputs. We formally define and evaluate these challenging scenarios. We fine-tune three LLMs and a state-of-the-art large vision-language model (LVLM) using RL with a representative policy-gradient algorithm and then test their performance on eight public datasets. Our results reveal that while RL fine-tuning improves baseline reasoning under idealized settings, performance declines significantly across all three non-ideal scenarios, exposing critical limitations in advanced reasoning capabilities. Although we propose a scenario-specific remediation method, our results suggest current methods leave these reasoning deficits largely unresolved. This work highlights that the reasoning abilities of large models are often overstated and underscores the importance of evaluating models under non-ideal scenarios. The code and data will be released at XXXX.

Chang Tian, Mingzhe Xing, Zenglin Shi et al.

Predicting web service traffic has significant social value, as it can be applied to various practical scenarios, including but not limited to dynamic resource scaling, load balancing, system anomaly detection, service-level agreement compliance, and fraud detection. Web service traffic is characterized by frequent and drastic fluctuations over time and are influenced by heterogeneous web user behaviors, making accurate prediction a challenging task. Previous research has extensively explored statistical approaches, and neural networks to mine features from preceding service traffic time series for prediction. However, these methods have largely overlooked the causal relationships between services. Drawing inspiration from causality in ecological systems, we empirically recognize the causal relationships between web services. To leverage these relationships for improved web service traffic prediction, we propose an effective neural network module, CCMPlus, designed to extract causal relationship features across services. This module can be seamlessly integrated with existing time series models to consistently enhance the performance of web service traffic predictions. We theoretically justify that the causal correlation matrix generated by the CCMPlus module captures causal relationships among services. Empirical results on real-world datasets from Microsoft Azure, Alibaba Group, and Ant Group confirm that our method surpasses state-of-the-art approaches in Mean Squared Error (MSE) and Mean Absolute Error (MAE) for predicting service traffic time series. These findings highlight the efficacy of leveraging causal relationships for improved predictions.

Sander Schildermans, Chang Tian, Ying Jiao et al.

Text-to-image (T2I) generation has advanced rapidly, yet faithfully capturing spatial relationships described in natural language prompts remains a major challenge. Prior efforts have addressed this issue through prompt optimization, spatially grounded generation, and semantic refinement. This work introduces a lightweight approach that augments prompts with tuple-based structured information, using a fine-tuned language model for automatic conversion and seamless integration into T2I pipelines. Experimental results demonstrate substantial improvements in spatial accuracy, without compromising overall image quality as measured by Inception Score. Furthermore, the automatically generated tuples exhibit quality comparable to human-crafted tuples. This structured information provides a practical and portable solution to enhance spatial relationships in T2I generation, addressing a key limitation of current large-scale generative systems.

Mingzhe Xing, Chang Tian, Jianan Zhang et al.

Modern large-scale networks introduce significant complexity in understanding network behaviors, increasing the risk of misconfiguration. Prior work proposed to understand network behaviors by mining network configurations, typically relying on domain-specific languages interfaced with formal models. While effective, they suffer from a steep learning curve and limited flexibility. In contrast, natural language (NL) offers a more accessible and interpretable interface, motivating recent research on NL-guided network behavior understanding. Recent advances in large language models (LLMs) further enhance this direction, leveraging their extensive prior knowledge of network concepts and strong reasoning capabilities. However, three key challenges remain: 1) numerous router devices with lengthy configuration files challenge LLM's long-context understanding ability; 2) heterogeneity across devices and protocols impedes scalability; and 3) complex network topologies and protocols demand advanced reasoning abilities beyond the current capabilities of LLMs. To tackle the above challenges, we propose NetMind, a novel framework for querying networks using NL. Our approach introduces a tree-based configuration chunking strategy to preserve semantic coherence while enabling efficient partitioning. We then construct a unified fact graph as an intermediate representation to normalize vendor-specific configurations. Finally, we design a hybrid imperative-declarative language to reduce the reasoning burden on LLMs and enhance precision. We contribute a benchmark consisting of NL question-answer pairs paired with network configurations. Experiments demonstrate that NetMind achieves accurate and scalable network behavior understanding, outperforming existing baselines.

Aleksa Jelaca, Ying Jiao, Chang Tian et al.

Text-to-image generation has advanced rapidly with large-scale multimodal training, yet fine-grained controllability remains a critical challenge. Counterfactual controllability, defined as the capacity to deliberately generate images that contradict common-sense patterns, remains a major challenge but plays a crucial role in enabling creativity and exploratory applications. In this work, we address this gap with a focus on counterfactual size (e.g., generating a tiny walrus beside a giant button) and propose an automatic prompt engineering framework that adapts base prompts into revised prompts for counterfactual images. The framework comprises three components: an image evaluator that guides dataset construction by identifying successful image generations, a supervised prompt rewriter that produces revised prompts, and a DPO-trained ranker that selects the optimal revised prompt. We construct the first counterfactual size text-image dataset and enhance the image evaluator by extending Grounded SAM with refinements, achieving a 114 percent improvement over its backbone. Experiments demonstrate that our method outperforms state-of-the-art baselines and ChatGPT-4o, establishing a foundation for future research on counterfactual controllability.

Chang Tian, Wenpeng Yin, Dan Li et al.

Few-shot named entity recognition (NER) systems recognize entities using a few labeled training examples. The general pipeline consists of a span detector to identify entity spans in text and an entity-type classifier to assign types to entities. Current span detectors rely on extensive manual labeling to guide training. Almost every span detector requires initial training on basic span features followed by adaptation to task-specific features. This process leads to repetitive training of the basic span features among span detectors. Additionally, metric-based entity-type classifiers, such as prototypical networks, typically employ a specific metric that gauges the distance between the query sample and entity-type referents, ultimately assigning the most probable entity type to the query sample. However, these classifiers encounter the sample dependency problem, primarily stemming from the limited samples available for each entity-type referent. To address these challenges, we proposed an improved few-shot NER pipeline. First, we introduce a steppingstone span detector that is pre-trained on open-domain Wikipedia data. It can be used to initialize the pipeline span detector to reduce the repetitive training of basic features. Second, we leverage a large language model (LLM) to set reliable entity-type referents, eliminating reliance on few-shot samples of each type. Our model exhibits superior performance with fewer training steps and human-labeled data compared with baselines, as demonstrated through extensive experiments on various datasets. Particularly in fine-grained few-shot NER settings, our model outperforms strong baselines, including ChatGPT. We will publicly release the code, datasets, LLM outputs, and model checkpoints.

Ke Ma, Yiliang Sang, Yang Ming et al.

Deep learning based channel state information (CSI) feedback in frequency division duplex systems has drawn much attention in both academia and industry. In this paper, we focus on integrating the Type-II codebook in the beyond fifth-generation (B5G) wireless systems with deep learning to enhance the performance of CSI feedback. In contrast to its counterpart in Release 16, the Type-II codebook in Release 17 (R17) exploits the angular-delay-domain partial reciprocity between uplink and downlink channels and selects part of angular-delay-domain ports for measuring and feeding back the downlink CSI, where the performance of the conventional deep learning methods is limited due to the deficiency of sparse structures. To address this issue, we propose the new paradigm of adopting deep learning to improve the performance of R17 Type-II codebook. Firstly, considering the relatively low signal-to-noise ratio of uplink channels, deep learning is utilized to refine the selection of the dominant angular-delay-domain ports, where the focal loss is harnessed to solve the class imbalance problem. Secondly, we propose to reconstruct the downlink CSI by way of deep learning based on the feedback of R17 Type-II codebook at the base station, where the information of sparse structures can be effectively leveraged. Finally, a weighted shortcut module is designed to facilitate the accurate reconstruction, and a two-stage loss function with the combination of the mean squared error and sum rate is proposed for adapting to actual multi-user scenarios. Simulation results demonstrate that our proposed angular-delay-domain port selection and CSI reconstruction paradigm can improve the sum rate performance by more than 10% compared with the traditional R17 Type-II codebook and deep learning benchmarks.

Chang Tian, An Liu, Guang Huang et al.

We propose a successive convex approximation based off-policy optimization (SCAOPO) algorithm to solve the general constrained reinforcement learning problem, which is formulated as a constrained Markov decision process (CMDP) in the context of average cost. The SCAOPO is based on solving a sequence of convex objective/feasibility optimization problems obtained by replacing the objective and constraint functions in the original problems with convex surrogate functions. At each iteration, the convex surrogate problem can be efficiently solved by Lagrange dual method even the policy is parameterized by a high-dimensional function. Moreover, the SCAOPO enables to reuse old experiences from previous updates, thereby significantly reducing the implementation cost when deployed in the real-world engineering systems that need to online learn the environment. In spite of the time-varying state distribution and the stochastic bias incurred by the off-policy learning, the SCAOPO with a feasible initial point can still provably converge to a Karush-Kuhn-Tucker (KKT) point of the original problem almost surely.