Selected work

Representative Papers

When Rules Learn: A Self-Evolving Agent for Legal Case Retrieval

Jun 15, 2026

Legal case retrieval remains highly challenging due to linguistic complexity and the need for precise lexical alignment, with BM25 still serving as a strong baseline. This work proposes a training-free, self-evolving framework that leverages large language model (LLM) agents within an automated evaluation environment to iteratively generate query rewriting rules, design validation experiments, and dynamically prune ineffective rules based on historical feedback, thereby optimizing BM25 performance. To the best of our knowledge, this is the first approach to endow rule-based query rewriting with self-evolution capabilities, effectively integrating LLMs’ prior knowledge with empirical experimental feedback. Evaluated on the LeCaRD-v2 Chinese legal retrieval benchmark, the method significantly outperforms non-evolutionary baselines—including handcrafted rules and greedy selection strategies—especially when powered by high-performance LLMs.

0 citationsRead paper

Extragradient Method for $(L_0, L_1)$-Lipschitz Root-finding Problems

Oct 25, 2025

Existing extragradient (EG) methods for minimax optimization, root-finding, and variational inequalities rely on the restrictive strong $L$-Lipschitz assumption, failing to capture complex operator structures arising in modern machine learning. Method: We propose the first adaptive step-size strategy based on dynamic operator-norm estimation, operating under the significantly milder $alpha$-symmetric $(L_0, L_1)$-Lipschitz condition. Contribution/Results: We establish sublinear convergence for monotone operators and linear convergence for strongly monotone operators. Moreover, we provide the first local convergence guarantee under the weak Minty condition—a substantially weaker requirement than standard monotonicity. Experiments demonstrate that our method achieves robust, practical convergence across diverse non-standard operator settings, markedly broadening the applicability of extragradient methods beyond classical assumptions.

0 citationsRead paper

Adaptive Twisting Sliding Control for Integrated Attack UAV's Autopilot and Guidance

Jan 17, 2025

Addressing the robust interception challenge of adversarial UAVs under complex environments and abrupt target maneuvers, this paper proposes an integrated guidance-and-control framework based on Adaptive Twisting-Structure Sliding Mode Control (ATSMC). Leveraging a two-dimensional coupled dynamics and relative kinematics model, we design a zero-effort miss–based sliding surface and pioneer the application of ATSMC to UAV integrated guidance-control architecture. The method requires no prior knowledge of target acceleration and simultaneously mitigates strong nonlinearities, model uncertainties, external disturbances, and sudden target direction changes. Simulation results demonstrate significantly improved interception accuracy, along with stable, rapid, and robust convergence even under high uncertainty and aggressive target maneuvers. This work establishes a novel paradigm for autonomous intelligent air combat interception.

0 citationsRead paper

Recent publications

Latest Papers

When Rules Learn: A Self-Evolving Agent for Legal Case Retrieval

Jun 15, 2026

Legal case retrieval remains highly challenging due to linguistic complexity and the need for precise lexical alignment, with BM25 still serving as a strong baseline. This work proposes a training-free, self-evolving framework that leverages large language model (LLM) agents within an automated evaluation environment to iteratively generate query rewriting rules, design validation experiments, and dynamically prune ineffective rules based on historical feedback, thereby optimizing BM25 performance. To the best of our knowledge, this is the first approach to endow rule-based query rewriting with self-evolution capabilities, effectively integrating LLMs’ prior knowledge with empirical experimental feedback. Evaluated on the LeCaRD-v2 Chinese legal retrieval benchmark, the method significantly outperforms non-evolutionary baselines—including handcrafted rules and greedy selection strategies—especially when powered by high-performance LLMs.

0 citationsRead paper

Extragradient Method for $(L_0, L_1)$-Lipschitz Root-finding Problems

Oct 25, 2025

Existing extragradient (EG) methods for minimax optimization, root-finding, and variational inequalities rely on the restrictive strong $L$-Lipschitz assumption, failing to capture complex operator structures arising in modern machine learning. Method: We propose the first adaptive step-size strategy based on dynamic operator-norm estimation, operating under the significantly milder $alpha$-symmetric $(L_0, L_1)$-Lipschitz condition. Contribution/Results: We establish sublinear convergence for monotone operators and linear convergence for strongly monotone operators. Moreover, we provide the first local convergence guarantee under the weak Minty condition—a substantially weaker requirement than standard monotonicity. Experiments demonstrate that our method achieves robust, practical convergence across diverse non-standard operator settings, markedly broadening the applicability of extragradient methods beyond classical assumptions.

0 citationsRead paper

Adaptive Twisting Sliding Control for Integrated Attack UAV's Autopilot and Guidance

Jan 17, 2025

Addressing the robust interception challenge of adversarial UAVs under complex environments and abrupt target maneuvers, this paper proposes an integrated guidance-and-control framework based on Adaptive Twisting-Structure Sliding Mode Control (ATSMC). Leveraging a two-dimensional coupled dynamics and relative kinematics model, we design a zero-effort miss–based sliding surface and pioneer the application of ATSMC to UAV integrated guidance-control architecture. The method requires no prior knowledge of target acceleration and simultaneously mitigates strong nonlinearities, model uncertainties, external disturbances, and sudden target direction changes. Simulation results demonstrate significantly improved interception accuracy, along with stable, rapid, and robust convergence even under high uncertainty and aggressive target maneuvers. This work establishes a novel paradigm for autonomous intelligent air combat interception.

0 citationsRead paper