arXiv subjects

Anni Li

Publications and source records attributed to Anni Li.

At least 19 recordsLinked to original sources

arXiv2026

Finite-time Convergent Control Barrier Functions with Feasibility Guarantees

This paper studies the problem of finite-time convergence to a prescribed safe set for nonlinear systems whose initial states violate the safety constraints. Existing Control Lyapunov-Barrier Functions (CLBFs) can enforce recovery to the safe set but may suffer from the issue of chattering and they do not explicitly consider control bounds. To address these limitations, we propose a new Control Barrier Function (CBF) formulation that guarantees finite-time convergence to the safe set while ensuring feasibility under control constraints. Specifically, we strengthen the initially violated safety constraint by introducing a parameter which enables the exploitation of the asymptotic property of a CBF to converge to the safe set in finite time. Furthermore, the conditions for the existence of such a CBF under control bounds to achieve finite-time convergence are derived via reachability analysis and constraint comparison, providing a systematic approach for parameter design. A case study on 2D obstacle avoidance is presented to demonstrate the effectiveness and advantages of the proposed method.

eess.SY↗

arXiv2026

Exact-Time Safety Recovery using Time-Varying Control Barrier Functions with Optimal Barrier Tracking

This paper is motivated by controllers developed for autonomous vehicles which occasionally result into conditions where safety is no longer guaranteed. We develop an exact-time safety recovery framework for any control-affine nonlinear system when its state is outside a safe region using time-varying Control Barrier Functions (CBFs) with optimal barrier tracking. Unlike conventional formulations that provide only conservative upper bounds on recovery time convergence, the proposed approach guarantees recovery to the safe set at a prescribed time. The key mechanism is an active barrier tracking condition that forces the barrier function to follow exactly a designer-specified recovery trajectory. This transforms safety recovery into a trajectory design problem. The recovery trajectory is parameterized and optimized to achieve optimal performance while preserving feasibility under input constraints, avoiding the aggressive corrective actions typically induced by conventional finite-time formulations. The safety recovery framework is applied to the roundabout traffic coordination problem for Connected and Automated Vehicles (CAVs), where any initially violated safe merging constraint is replaced by an exact-time recovery barrier constraint to ensure safety guarantee restoration before CAV conflict points are reached. Simulation results demonstrate improved feasibility and performance.

eess.SY↗

arXiv2026

Robust Dynamic Pricing and Admission Control with Fairness Guarantees

Dynamic pricing is commonly used to regulate congestion in shared service systems. This paper is motivated by the fact that in the presence of users with varying price sensitivity (responsiveness), conventional monotonic pricing can lead to unfair outcomes by disproportionately excluding price-elastic users, particularly under high or uncertain demand. We therefore develop a fairness-oriented mechanism under demand uncertainty. The paper's contributions are twofold. First, we show that when fairness is imposed as a hard state constraint, the optimal (revenue maximizing) pricing policy is generally non-monotonic in demand. This structural result departs fundamentally from standard surge pricing rules and reveals that price reduction under heavy load may be necessary to maintain equitable access. Second, we address the problem that price elasticity among heterogeneous users is unobservable. To solve it, we develop a robust dynamic pricing and admission control framework that enforces capacity and fairness constraints for all user type distributions consistent with aggregate measurements. By integrating integral High Order Control Barrier Functions (iHOCBFs) with a robust optimization framework under uncertain user-type distribution, we obtain a controller that guarantees forward invariance of safety and fairness constraints while optimizing revenue. Numerical experiments demonstrate improved fairness and revenue performance relative to monotonic surge pricing policies.

eess.SY↗

arXiv2026

Robust Taylor-Lagrange Control for Safety-Critical Systems

Solving safety-critical control problem has widely adopted the Control Barrier Function (CBF) method. However, the existence of a CBF is only a sufficient condition for system safety. The recently proposed Taylor-Lagrange Control (TLC) method addresses this limitation, but is vulnerable to the feasibility preservation problem (e.g., inter-sampling effect). In this paper, we propose a robust TLC (rTLC) method to address the feasibility preservation problem. Specifically, the rTLC method expands the safety function at an order higher than the relative degree of the function using Taylor's expansion with Lagrange remainder, which allows the control to explicitly show up at the current time instead of the future time in the TLC method. The rTLC method naturally addresses the feasibility preservation problem with only one hyper-parameter (the discretization time interval size during implementation), which is much less than its counterparts. Finally, we illustrate the effectiveness of the proposed rTLC method through an adaptive cruise control problem, and compare it with existing safety-critical control methods.

eess.SY↗

arXiv2025

Taylor-Lagrange Control for Safety-Critical Systems

This paper proposes a novel Taylor-Lagrange Control (TLC) method for nonlinear control systems to ensure the safety and stability through Taylor's theorem with Lagrange remainder. To achieve this, we expand a safety or stability function with respect to time along the system dynamics using the Lie derivative and Taylor's theorem. This expansion enables the control input to appear in the Taylor series at an order equivalent to the relative degree of the function. We show that the proposed TLC provides necessary and sufficient conditions for system safety and is applicable to systems and constraints of arbitrary relative degree. The TLC exhibits connections with existing Control Barrier Function (CBF) and Control Lyapunov Function (CLF) methods, and it further extends the CBF and CLF methods to the complex domain, especially for higher order cases. Compared to High-Order CBFs (HOCBFs), TLC is less restrictive as it does not require forward invariance of the intersection of a set of safe sets while HOCBFs do. We employ TLC to reformulate a constrained optimal control problem as a sequence of quadratic programs with a zero-order hold implementation method, and demonstrate the safety of zero-order hold TLC using an event-triggered control method to address inter-sampling effects. Finally, we illustrate the effectiveness of the proposed TLC method through an adaptive cruise control system and a robot control problem, and compare it with existing CBF methods.

eess.SY↗

arXiv2025

Thinkquel: A Model Dedicated to Text-to-dbt Using Synthetic Data and a Span-Aware Objective

Transforming natural-language requests into reliable, production-ready data transformations remains challenging: correctness depends on precise schema linking and warehouse-specific SQL dialects, while the strongest supervision available during training--execution success and result matching--are provided only at the sequence level. At the same time, assembling large, execution-validated corpora is costly, and token-level objectives misalign with these global signals, yielding unstable optimization and limited portability. We introduce Thinkquel, a fine-tuned model for producing robust, portable, and execution-validated database queries. Methodologies in Thinkquel integrates a novel synthetic data pipeline, TS-SQL, that leverages dbt as a portable intermediate representation with a span-aware reinforcement learning objective, and Token-Sequence GRPO (TS-GRPO), specifically designed to bridge the gap between token-level training signals and sequence-level execution rewards when finetuning LLMs. On the 500-example TS-SQL test set, Thinkquel (32B) reaches 93.2% execution success and 61.8% exact-result match with a two-stage SFT curriculum, improving over the base model by 67.2% (exec.) and 44.4% (match). In Spider (14B) experiments, TS-GRPO increases training stability and speeds convergence of the execution-match reward relative to GRPO and GSPO.

cs.AI↗

arXiv2025

LAWCAT: Efficient Distillation from Quadratic to Linear Attention with Convolution across Tokens for Long Context Modeling

Although transformer architectures have achieved state-of-the-art performance across diverse domains, their quadratic computational complexity with respect to sequence length remains a significant bottleneck, particularly for latency-sensitive long-context applications. While recent linear-complexity alternatives are increasingly powerful, effectively training them from scratch is still resource-intensive. To overcome these limitations, we propose LAWCAT (Linear Attention with Convolution Across Time), a novel linearization framework designed to efficiently transfer the capabilities of pre-trained transformers into a performant linear attention architecture. LAWCAT integrates causal Conv1D layers to enhance local dependency modeling and employs normalized gated linear attention to improve generalization across varying context lengths. Our comprehensive evaluations demonstrate that, distilling Mistral-7B with only 1K-length sequences yields over 90\% passkey retrieval accuracy up to 22K tokens, significantly extending its effective context window. Similarly, Llama3.2-1B LAWCAT variant achieves competitive performance on S-NIAH 1\&2\&3 tasks (1K-8K context length) and BABILong benchmark (QA2\&QA3, 0K-16K context length), requiring less than 0.1\% pre-training tokens compared with pre-training models. Furthermore, LAWCAT exhibits faster prefill speeds than FlashAttention-2 for sequences exceeding 8K tokens. LAWCAT thus provides an efficient pathway to high-performance, long-context linear models suitable for edge deployment, reducing reliance on extensive long-sequence training data and computational resources. Code is released at: https://github.com/zeyuliu1037/LAWCAT

cs.CL↗

arXiv2025

A Cooperative Compliance Control Framework for Socially Optimal Mixed Traffic Routing

In mixed traffic environments, where Connected and Autonomed Vehicles (CAVs) coexist with potentially non-cooperative Human-Driven Vehicles (HDVs), the self-centered behavior of human drivers may compromise the efficiency, optimality, and safety of the overall traffic network. In this paper, we propose a Cooperative Compliance Control (CCC) framework for mixed traffic routing, where a Social Planner (SP) optimizes vehicle routes for system-wide optimality while a compliance controller incentivizes human drivers to align their behavior with route guidance from the SP through a "refundable toll" scheme. A key challenge arises from the heterogeneous and unknown response models of different human driver types to these tolls, making it difficult to design a proper controller and achieve desired compliance probabilities over the traffic network. To address this challenge, we employ Control Lyapunov Functions (CLFs) to adaptively correct (learn) crucial components of our compliance probability model online, construct data-driven feedback controllers, and demonstrate that we can achieve the desired compliance probability for HDVs, thereby contributing to the social optimality of the traffic network.

eess.SY↗

arXiv2025

Routing Guidance for Emerging Transportation Systems with Improved Dynamic Trip Equity

This paper presents a dynamic routing guidance system that optimizes route recommendations for individual vehicles in an emerging transportation system while enhancing travelers' trip equity. We develop a framework to quantify trip quality and equity in dynamic travel environments, providing new insights into how routing guidance influences equity in road transportation. Our approach enables real-time routing by incorporating both monitored and anticipated traffic congestion. We provide conditions that ensure perfect trip equity for all travelers in a free-flow network. Simulation studies on 1,000 vehicles traversing an urban road network in Boston demonstrate that our method improves trip equity by approximately 11.4\% compared to the shortest-route strategy. In addition, the results reveal that our approach redistributes travel costs across vehicle types through route optimization, contributing to a more equitable transportation system.

eess.SY↗

arXiv2024

0.7 MW Yb:YAG pumped degenerate optical parametric oscillator at 2.06 {\mu}m

Frequency comb and field-resolved broadband absorption spectroscopy are promising techniques for rapid, precise, and sensitive detection of short-lived atmospheric pollutants on-site. Enhancing detection sensitivity in absorption spectroscopy hinges on bright sources that cover molecular resonances and fast signal modulation techniques to implement lock-in detection schemes efficiently. Yb:YAG thin-disk lasers, combined with optical parametric oscillators (OPO), present a compelling solution to fulfill these requirements. In this work, we report on a bright OPO pumped by a Yb:YAG thin-disk Kerr-lens mode-locked oscillator delivering 2.8 W, 114 fs pulses at 2.06 {\mu}m with an averaged energy of 90 nJ. The OPO cavity operates at 30.9 MHz pulse repetition rates, the second harmonic of the pump cavity, allowing for broadband, efficient, and dispersion-free modulation of the OPO output pulses at 15.45 MHz rate. With 13% optical-to-optical conversion efficiency and a high-frequency intra-cavity modulation, this scalable scheme holds promise to advance the detection sensitivity and frontiers of field-resolved spectroscopic techniques.

physics.optics↗

arXiv2024

Towards Achieving Cooperation Compliance of Human Drivers in Mixed Traffic

We consider a mixed-traffic environment in transportation systems, where Connected and Automated Vehicles (CAVs) coexist with potentially non-cooperative Human-Driven Vehicles (HDVs). We develop a cooperation compliance control framework to incentivize HDVs to align their behavior with socially optimal objectives using a ``refundable toll'' scheme so as to achieve a desired compliance probability for all non-compliant HDVs through a feedback control mechanism combining global with local (individual) components. We apply this scheme to the lane-changing problem, where a ``Social Planner'' provides references to the HDVs, measures their state errors, and induces cooperation compliance for safe lane-changing through a refundable toll approach. Simulation results are included to show the effectiveness of our cooperation compliance controller in terms of improved compliance and lane-changing maneuver safety and efficiency when non-cooperative HDVs are present.

eess.SY↗

arXiv2024

AFLoRA: Adaptive Freezing of Low Rank Adaptation in Parameter Efficient Fine-Tuning of Large Models

We present a novel Parameter-Efficient Fine-Tuning (PEFT) method, dubbed as Adaptive Freezing of Low Rank Adaptation (AFLoRA). Specifically, for each pre-trained frozen weight tensor, we add a parallel path of trainable low-rank matrices, namely a down-projection and an up-projection matrix, each of which is followed by a feature transformation vector. Based on a novel freezing score, we the incrementally freeze these projection matrices during fine-tuning to reduce the computation and alleviate over-fitting. Our experimental results demonstrate that we can achieve state-of-the-art performance with an average improvement of up to $0.85\%$ as evaluated on GLUE benchmark while yeilding up to $9.5\times$ fewer average trainable parameters. While compared in terms of runtime, AFLoRA can yield up to $1.86\times$ improvement as opposed to similar PEFT alternatives. Besides the practical utility of our approach, we provide insights on the trainability requirements of LoRA paths at different modules and the freezing schedule for the different projection matrices. Code will be released.

cs.CL↗

arXiv2024

Robust Optimal Lane-changing Control for Connected Autonomous Vehicles in Mixed Traffic

We derive time and energy-optimal policies for a Connected Autonomous Vehicle (CAV) to execute lane change maneuvers in mixed traffic, i.e., in the presence of both CAVs and Human Driven Vehicles (HDVs). These policies are also shown to be robust with respect to the unpredictable behavior of HDVs by exploiting CAV cooperation which can eliminate or greatly reduce the interaction between CAVs and HDVs. We derive a simple threshold-based criterion on the initial relative distance between two cooperating CAVs based on which an optimal policy is selected such that the lane-changing CAV merges ahead of a cooperating CAV in the target lane; in this case, the lane-changing CAV's trajectory becomes independent of HDV behavior. Otherwise, the interaction between CAVs and neighboring HDVs is formulated as a bilevel optimization problem with an appropriate behavioral model for an HDV, and an iterated best response (IBR) method is used to determine an equilibrium. We demonstrate the convergence of the IBR process under certain conditions. Furthermore, Control Barrier Functions (CBFs) are implemented to ensure the robustness of lane-changing behaviors by guaranteeing safety in both longitudinal and lateral directions despite HDV disturbances. Simulation results validate the effectiveness of our CAV controllers in terms of cost, safety guarantees, and limited disruption to traffic flow. Additionally, we demonstrate the robustness of the lane-changing behaviors in the presence of uncontrollable HDVs.

eess.SY↗

arXiv2024

LMUFormer: Low Complexity Yet Powerful Spiking Model With Legendre Memory Units

Transformer models have demonstrated high accuracy in numerous applications but have high complexity and lack sequential processing capability making them ill-suited for many streaming applications at the edge where devices are heavily resource-constrained. Thus motivated, many researchers have proposed reformulating the transformer models as RNN modules which modify the self-attention computation with explicit states. However, these approaches often incur significant performance degradation. The ultimate goal is to develop a model that has the following properties: parallel training, streaming and low-cost inference, and SOTA performance. In this paper, we propose a new direction to achieve this goal. We show how architectural modifications to a recurrent model can help push its performance toward Transformer models while retaining its sequential processing capability. Specifically, inspired by the recent success of Legendre Memory Units (LMU) in sequence learning tasks, we propose LMUFormer, which augments the LMU with convolutional patch embedding and convolutional channel mixer. Moreover, we present a spiking version of this architecture, which introduces the benefit of states within the patch embedding and channel mixer modules while simultaneously reducing the computing complexity. We evaluated our architectures on multiple sequence datasets. In comparison to SOTA transformer-based models within the ANN domain on the SCv2 dataset, our LMUFormer demonstrates comparable performance while necessitating a remarkable 53 times reduction in parameters and a substantial 65 times decrement in FLOPs. Additionally, owing to our model's proficiency in real-time data processing, we can achieve a 32.03% reduction in sequence length, all while incurring an inconsequential decline in performance. Our code is publicly available at https://github.com/zeyuliu1037/LMUFormer.git.

cs.NE↗

arXiv2023

Induced Generative Adversarial Particle Transformers

In high energy physics (HEP), machine learning methods have emerged as an effective way to accurately simulate particle collisions at the Large Hadron Collider (LHC). The message-passing generative adversarial network (MPGAN) was the first model to simulate collisions as point, or ``particle'', clouds, with state-of-the-art results, but suffered from quadratic time complexity. Recently, generative adversarial particle transformers (GAPTs) were introduced to address this drawback; however, results did not surpass MPGAN. We introduce induced GAPT (iGAPT) which, by integrating ``induced particle-attention blocks'' and conditioning on global jet attributes, not only offers linear time complexity but is also able to capture intricate jet substructure, surpassing MPGAN in many metrics. Our experiments demonstrate the potential of iGAPT to simulate complex HEP data accurately and efficiently.

hep-ex↗

arXiv2023

Spiking Neural Networks with Dynamic Time Steps for Vision Transformers

Spiking Neural Networks (SNNs) have emerged as a popular spatio-temporal computing paradigm for complex vision tasks. Recently proposed SNN training algorithms have significantly reduced the number of time steps (down to 1) for improved latency and energy efficiency, however, they target only convolutional neural networks (CNN). These algorithms, when applied on the recently spotlighted vision transformers (ViT), either require a large number of time steps or fail to converge. Based on analysis of the histograms of the ANN and SNN activation maps, we hypothesize that each ViT block has a different sensitivity to the number of time steps. We propose a novel training framework that dynamically allocates the number of time steps to each ViT module depending on a trainable score assigned to each timestep. In particular, we generate a scalar binary time step mask that filters spikes emitted by each neuron in a leaky-integrate-and-fire (LIF) layer. The resulting SNNs have high activation sparsity and require only accumulate operations (AC), except for the input embedding layer, in contrast to expensive multiply-and-accumulates (MAC) needed in traditional ViTs. This yields significant improvements in energy efficiency. We evaluate our training framework and resulting SNNs on image recognition tasks including CIFAR10, CIFAR100, and ImageNet with different ViT architectures. We obtain a test accuracy of 95.97% with 4.97 time steps with direct encoding on CIFAR10.

cs.CV↗

arXiv2023

Safe Optimal Interactions Between Automated and Human-Driven Vehicles in Mixed Traffic with Event-triggered Control Barrier Functions

This paper studies safe driving interactions between Human-Driven Vehicles (HDVs) and Connected and Automated Vehicles (CAVs) in mixed traffic where the dynamics and control policies of HDVs are unknown and hard to predict. In order to address this challenge, we employ event-triggered Control Barrier Functions (CBFs) to estimate the HDV model online, construct data-driven and state-feedback safety controllers, and transform constrained optimal control problems for CAVs into a sequence of event-triggered quadratic programs. We show that we can ensure collision-free between HDVs and CAVs and demonstrate the robustness and flexibility of our framework on different types of human drivers in lane-changing scenarios while guaranteeing safety with human-in-the-loop interactions.

eess.SY↗

arXiv2023

Maximizing Safety and Efficiency for Cooperative Lane-Changing: A Minimally Disruptive Approach

This paper addresses cooperative lane-changing maneuvers in mixed traffic, aiming to minimize traffic flow disruptions while accounting for uncooperative vehicles. The proposed approach adopts controllers combining Optimal control with Control Barrier Functions (OCBF controllers) which guarantee spatio-temporal constraints through the use of fixed-time convergence. Additionally, we introduce robustness to disturbances by deriving a method for handling worst-case disturbances using the dual of a linear programming problem. We present a near-optimal solution that ensures safety, optimality, and robustness to changing behavior of uncooperative vehicles. Simulations demonstrate the effectiveness of the proposed approach in enhancing efficiency and safety.

eess.SY↗