LCLu Cheng

Lu Cheng

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article

Previous Organizations / Institutes

The organizations / institutes from the scholar's other articles in the past ten years


Lu Cheng has a total of 173 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    Graph-augmented LLMs for social media-based clinical trial recruitment

    by Xiaofan Zhou, Zisu Wang, Janice Krieger, Mohan Zalake, Lu Cheng

  • Open Access

    The Confidence Trap: Calibration Attacks for Graph Neural Networks

    by Cuong Dang, Jiahao Zhang, Hieu Ta Quang, Dung Le, Lu Cheng, Suhang Wang

  • Open Access

    SafeCRS: Personalized Safety Alignment for LLM-Based Conversational Recommender Systems

    by Haochang Hao, Yifan Xu, Xinzhuo Li, Yingqiang Ge, Lu Cheng

  • Open Access

    ConfDock: Atom-specific Uncertainty Quantification for Molecular Docking via Conformal Prediction

    by Haochang Hao, Nour Elhendawy, Yihang Wang, Cheng Lu

    Molecular docking is widely used in structure-based drug discovery, yet most approaches provide point estimates without rigorous uncertainty quantification. This limitation makes it difficult to assess when a predicted pose should be trusted, especially when docking methods are applied to diverse protein-ligand systems. We present ConfDock, a conformal prediction (CP) framework for constructing atom-specific prediction intervals for ligand docking poses. ConfDock combines graph neural network (GNN) based quantile estimation with split conformal calibration, producing intervals that adapt to local protein-ligand environments while retaining distribution-free finite-sample coverage guarantees. We evaluate ConfDock on 238 protein-ligand complexes across four docking methods representing distinct computational paradigms. The proposed approach yields substantially narrower prediction intervals compared to standard split CP (57.2% average reduction in mean interval width, up to 74.5%) while maintaining target coverage across all evaluated settings. Ablation analysis indicates that the GNN captures the dominant structure-dependent variability in uncertainty, whereas the conformal calibration step provides a bounded adjustment to ensure coverage guarantees. These results demonstrate that combining learned, structure-aware quantile estimation with conformal calibration enables rigorous uncertainty quantification for molecular docking at atom-level resolution.

  • Open Access

    Graph-Augmented LLMs for Social-Media–based Clinical Trial Recruitment

    by Xiaofan Zhou, Zisu Wang, Janice Krieger, Mohan Zalake, Lu Cheng

    ObjectiveClinical trials (CTs) are essential for advancing disease diagnosis and treatment; however, recruiting eligible participants remains a major bottleneck. Existing recruitment strategies, such as advertisements and electronic health record–based screening, are often time-consuming and inefficient. Social networks provide a promising alternative source for identifying potential participants, but this direction remains under-explored due to sparse user information and the limitations of current methods.MethodsTo address these challenges, we propose a novel framework to directly predict whether a social media user is a potential candidate for CTs. To mitigate the issue of sparse user information, we introduce a summarization-based method to enhance prediction quality. Furthermore, motivated by the observation that users may share similar behaviors under specific eligibility criteria, we design a two-stage graph-based method to improve prediction performance further.ResultsExperiments conducted on a Reddit CT dataset demonstrate that our proposed framework significantly outperforms existing methods. Ablation studies further confirm that both the summarization module and the two-stage graph-based method contribute to the observed performance improvements.ConclusionOur approach substantially improves the accuracy of identifying potential CT participants by integrating summarization and two-stage graph-based modeling. These results highlight the strong potential of leveraging social media platforms for scalable and efficient CT recruitment.

  • Open Access

    Latent Thoughts Tuning: Bridging Context and Reasoning with Fused Information in Latent Tokens

    by Weihao Liu, Dehai Min, Lu Cheng

    While explicit Chain-of-Thought (CoT) equips Large Language Models (LLMs) with strong reasoning capabilities, it requires models to verbalize every intermediate step in text tokens, constraining the model thoughts to the discrete vocabulary space. Recently, reasoning in continuous latent space has emerged as a promising alternative, enabling more robust inference and flexible computation beyond discrete token constraints. However, current latent paradigms often suffer from feature collapse and instability, stemming from distribution mismatches when recurrently using hidden states as the input embeddings, or alignment issues when relying on assistant models. To address this, we propose Latent Thoughts Tuning (LT-Tuning), a framework that redefines how latent thoughts are constructed and deployed. Instead of relying solely on raw hidden states, our method introduces a Context-Prediction-Fusion mechanism that jointly leveraging contextual hidden states and predictive semantic guidance from the vocabulary embedding space. Combined with a progressive three-stage curriculum learning pipeline, LT-Tuning also enables dynamically switching between latent and explicit thinking modes. Experiments demonstrate that our method outperforms existing latent reasoning baselines, effectively mitigating feature collapse and achieving robust reasoning accuracy.

  • Open Access

    Evaluating Social Bias in RAG Systems: When External Context Helps and Reasoning Hurts

    by Shweta Parihar, Lu Cheng

    Social biases inherent in large language models (LLMs) raise significant fairness concerns. Retrieval-Augmented Generation (RAG) architectures, which retrieve external knowledge sources to enhance the generative capabilities of LLMs, remain susceptible to the same bias-related challenges. This work focuses on evaluating and understanding the social bias implications of RAG. Through extensive experiments across various retrieval corpora, LLMs, and bias evaluation datasets, encompassing more than 13 different bias types, we surprisingly observe a reduction in bias in RAG. This suggests that the inclusion of external context can help counteract stereotype-driven predictions, potentially improving fairness by diversifying the contextual grounding of the model's outputs. To better understand this phenomenon, we then explore the model's reasoning process by integrating Chain-of-Thought (CoT) prompting into RAG while assessing the faithfulness of the model's CoT. Our experiments reveal that the model's bias inclinations shift between stereotype and anti-stereotype responses as more contextual information is incorporated from the retrieved documents. Interestingly, we find that while CoT enhances accuracy, contrary to the bias reduction observed with RAG, it increases overall bias across datasets, highlighting the need for bias-aware reasoning frameworks that can mitigate this trade-off.

  • Open Access

    Context-Aware Counterfactual Data Augmentation for Gender Bias Mitigation in Language Models

    by Shweta Parihar, Liu Guangliang, Natalie Parde, Lu Cheng

    A challenge in mitigating social bias in fine-tuned language models (LMs) is the potential reduction in language modeling capability, which can harm downstream performance. Counterfactual data augmentation (CDA), a widely used method for fine-tuning, highlights this issue by generating synthetic data that may align poorly with real-world distributions or creating overly simplistic counterfactuals that ignore the social context of altered sensitive attributes (e.g., gender) in the pretraining corpus. To address these limitations, we propose a simple yet effective context-augmented CDA method, Context-CDA, which uses large LMs to enhance the diversity and contextual relevance of the debiasing corpus. By minimizing discrepancies between the debiasing corpus and pretraining data through augmented context, this approach ensures better alignment, enhancing language modeling capability. We then employ uncertainty-based filtering to exclude generated counterfactuals considered low-quality by the target smaller LMs (i.e., LMs to be debiased), further improving the fine-tuning corpus quality. Experimental results on gender bias benchmarks demonstrate that Context-CDA effectively mitigates bias without sacrificing language modeling performance while offering insights into social biases by analyzing distribution shifts in next-token generation probabilities.