2026
Forward-Chaining Temporal Point Process
Chao Yang, Wendi Ren, Shuang Li
The International Conference on Machine Learning (ICML) 2026
FC-TPP enables controllable and constraint-aware continuous-time event sequence generation by coupling temporal point processes with differentiable multi-hop logical reasoning in latent symbolic space.
Forward-Chaining Temporal Point Process
Chao Yang, Wendi Ren, Shuang Li
The International Conference on Machine Learning (ICML) 2026
FC-TPP enables controllable and constraint-aware continuous-time event sequence generation by coupling temporal point processes with differentiable multi-hop logical reasoning in latent symbolic space.
Branching Diffusion for Point Processes in Time and Space
Chao Yang, Wenjie Shen, Shuang Li
The International Conference on Machine Learning (ICML) 2026
We propose a non-autoregressive diffusion framework for temporal and spatio-temporal point processes derived from the Wasserstein–Fisher–Rao gradient flow, enabling principled generation via configuration-level denoising.
Branching Diffusion for Point Processes in Time and Space
Chao Yang, Wenjie Shen, Shuang Li
The International Conference on Machine Learning (ICML) 2026
We propose a non-autoregressive diffusion framework for temporal and spatio-temporal point processes derived from the Wasserstein–Fisher–Rao gradient flow, enabling principled generation via configuration-level denoising.
Deliberate-When-Needed: Flow-Reasoner for Neuro-Symbolic Continuous Thought
Wenjie Shen*, Boyang Li*, Chao Yang, Shuang Li (* equal contribution)
The International Conference on Artificial Intelligence and Statistics (AISTATS) 2026
We propose Deliberate-When-Needed, a continuous-time model that marries neural ODEs with temporal point processes to explain action generation through interpretable, multi-hop reasoning.
Deliberate-When-Needed: Flow-Reasoner for Neuro-Symbolic Continuous Thought
Wenjie Shen*, Boyang Li*, Chao Yang, Shuang Li (* equal contribution)
The International Conference on Artificial Intelligence and Statistics (AISTATS) 2026
We propose Deliberate-When-Needed, a continuous-time model that marries neural ODEs with temporal point processes to explain action generation through interpretable, multi-hop reasoning.
From Counts to Preferences: Preference-Driven Models for Spatio-Temporal Event Data
Chao Yang*, Yiling Kuang*, Shuang Li (* equal contribution)
The International Conference on Artificial Intelligence and Statistics (AISTATS) 2026
We introduce a preference-driven framework that models event distributions through a two-stage “consider–then–choose” process.
From Counts to Preferences: Preference-Driven Models for Spatio-Temporal Event Data
Chao Yang*, Yiling Kuang*, Shuang Li (* equal contribution)
The International Conference on Artificial Intelligence and Statistics (AISTATS) 2026
We introduce a preference-driven framework that models event distributions through a two-stage “consider–then–choose” process.
Learning Human Habits with Rule-Guided Active Inference
Zhiren Gong, Chao Yang, Wendi Ren, Shuang Li
The International Conference on Learning Representations (ICLR) 2026
A novel cognitive framework in which agents iteratively update their internal world models based on historical data through optimization, and use these models to plan and select future actions.
Learning Human Habits with Rule-Guided Active Inference
Zhiren Gong, Chao Yang, Wendi Ren, Shuang Li
The International Conference on Learning Representations (ICLR) 2026
A novel cognitive framework in which agents iteratively update their internal world models based on historical data through optimization, and use these models to plan and select future actions.
2025
Discovering Logic-Informed Intrinsic Rewards to Explain Human Policies
Chengzhi Cao*, Yinghao Fu*, Chao Yang, Shuang Li (* equal contribution)
ICML Workshop (PRAL) 2025
Uses inverse reinforcement learning with a neural logic tree generator to extract interpretable logical rules and intrinsic rewards from expert demonstrations, explaining human decision policies.
Discovering Logic-Informed Intrinsic Rewards to Explain Human Policies
Chengzhi Cao*, Yinghao Fu*, Chao Yang, Shuang Li (* equal contribution)
ICML Workshop (PRAL) 2025
Uses inverse reinforcement learning with a neural logic tree generator to extract interpretable logical rules and intrinsic rewards from expert demonstrations, explaining human decision policies.
Evolving Minds: Logic-Informed Inference from Temporal Action Patterns
Chao Yang, Shuting Cui, Yang Yang, Shuang Li
The International Conference on Machine Learning (ICML) 2025
Models latent human mental states from irregular actions using a logic-informed TPP with variational EM, enabling accurate inference and prediction of behavior.
Evolving Minds: Logic-Informed Inference from Temporal Action Patterns
Chao Yang, Shuting Cui, Yang Yang, Shuang Li
The International Conference on Machine Learning (ICML) 2025
Models latent human mental states from irregular actions using a logic-informed TPP with variational EM, enabling accurate inference and prediction of behavior.
Flow-Based Delayed Hawkes Process
Chao Yang, Wendi Ren, Shuang Li
The Conference on Uncertainty in Artificial Intelligence (UAI) 2025
Enhances delayed Hawkes processes with normalizing flows to achieve greater flexibility and expressiveness while preserving interpretability, supported by theoretical guarantees and improved performance.
Flow-Based Delayed Hawkes Process
Chao Yang, Wendi Ren, Shuang Li
The Conference on Uncertainty in Artificial Intelligence (UAI) 2025
Enhances delayed Hawkes processes with normalizing flows to achieve greater flexibility and expressiveness while preserving interpretability, supported by theoretical guarantees and improved performance.
Who Should Be Consulted? Targeted Expert Selection for Rare Disease Diagnosis
Yinghao Fu*, Chao Yang*, Xinye Chen, Yuting Yan, Shuang Li (* equal contribution)
ICML Workshop (CFAgentic) 2025 Oral
Proposes a dynamic, rule-guided framework that selectively engages the most relevant experts for each case, improving diagnostic accuracy and efficiency in rare disease decision-making.
Who Should Be Consulted? Targeted Expert Selection for Rare Disease Diagnosis
Yinghao Fu*, Chao Yang*, Xinye Chen, Yuting Yan, Shuang Li (* equal contribution)
ICML Workshop (CFAgentic) 2025 Oral
Proposes a dynamic, rule-guided framework that selectively engages the most relevant experts for each case, improving diagnostic accuracy and efficiency in rare disease decision-making.
2024
Neuro-Symbolic Temporal Point Processes
Yang Yang, Chao Yang, Boyang Li, Yinghao Fu, Shuang Li
The International Conference on Machine Learning (ICML) 2024
Introduces a neuro-symbolic TPP framework that learns compact, interpretable temporal logic rules end-to-end, achieving efficient and accurate explanation of irregular events.
Neuro-Symbolic Temporal Point Processes
Yang Yang, Chao Yang, Boyang Li, Yinghao Fu, Shuang Li
The International Conference on Machine Learning (ICML) 2024
Introduces a neuro-symbolic TPP framework that learns compact, interpretable temporal logic rules end-to-end, achieving efficient and accurate explanation of irregular events.
Latent Logic Tree Extraction for Event Sequence Explanation from LLMs
Zitao Song, Chao Yang, Chaojie Wang, Bo An, Shuang Li
The International Conference on Machine Learning (ICML) 2024
Combines LLM priors with a TPP-based EM framework to extract interpretable logic-tree explanations for event sequences, using GFlowNet to efficiently explore complex rule spaces.
Latent Logic Tree Extraction for Event Sequence Explanation from LLMs
Zitao Song, Chao Yang, Chaojie Wang, Bo An, Shuang Li
The International Conference on Machine Learning (ICML) 2024
Combines LLM priors with a TPP-based EM framework to extract interpretable logic-tree explanations for event sequences, using GFlowNet to efficiently explore complex rule spaces.
Unveiling Latent Causal Rules: A Temporal Point Process Approach for Abnormal Event Explanation
Yiling Kuang*, Chao Yang*, Yang Yang, Shuang Li (* equal contribution)
The International Conference on Artificial Intelligence and Statistics (AISTATS) 2024
We use temporal point processes and an EM framework to uncover latent causal “if–then” rules, enabling interpretable explanations and root-cause analysis of abnormal events.
Unveiling Latent Causal Rules: A Temporal Point Process Approach for Abnormal Event Explanation
Yiling Kuang*, Chao Yang*, Yang Yang, Shuang Li (* equal contribution)
The International Conference on Artificial Intelligence and Statistics (AISTATS) 2024
We use temporal point processes and an EM framework to uncover latent causal “if–then” rules, enabling interpretable explanations and root-cause analysis of abnormal events.
Discovering Intrinsic Spatial-Temporal Logic Rules to Explain Human Actions
Chengzhi Cao, Chao Yang, Ruimao Zhang, Shuang Li
The Conference on Neural Information Processing Systems (NeurIPS) 2024
We automatically discover interpretable spatial-temporal logic rules from trajectory data using an EM framework, enabling explainable and accurate modeling of human actions.
Discovering Intrinsic Spatial-Temporal Logic Rules to Explain Human Actions
Chengzhi Cao, Chao Yang, Ruimao Zhang, Shuang Li
The Conference on Neural Information Processing Systems (NeurIPS) 2024
We automatically discover interpretable spatial-temporal logic rules from trajectory data using an EM framework, enabling explainable and accurate modeling of human actions.
2023
Counterfactual Optimization of Treatment Policies Based on Temporal Point Process
Zilin Jing, Chao Yang, Shuang Li
ICML Workshop (IMLH) 2023
We propose a temporal point process-based framework to optimize specific treatment actions via counterfactual analysis, providing targeted, interpretable feedback under medical constraints.
Counterfactual Optimization of Treatment Policies Based on Temporal Point Process
Zilin Jing, Chao Yang, Shuang Li
ICML Workshop (IMLH) 2023
We propose a temporal point process-based framework to optimize specific treatment actions via counterfactual analysis, providing targeted, interpretable feedback under medical constraints.
2022
Reinforcement Logic Rule Learning for Temporal Point Processes
Chao Yang, Lu Wang, Kun Gao, Shuang Li
ICML Workshop (IMLH) 2022
We learn interpretable temporal logic rules for point processes by iteratively expanding a rule set, using reinforcement learning to efficiently search complex rule spaces and improve event likelihood.
Reinforcement Logic Rule Learning for Temporal Point Processes
Chao Yang, Lu Wang, Kun Gao, Shuang Li
ICML Workshop (IMLH) 2022
We learn interpretable temporal logic rules for point processes by iteratively expanding a rule set, using reinforcement learning to efficiently search complex rule spaces and improve event likelihood.