arXiv subjects

Mengqi Li

Publications and source records attributed to Mengqi Li.

10 recordsLinked to original sources

arXiv2026

Parity Selection Rule for Information and Dissipation in Driven Steady States

Tight equalities between symmetric information and entropy production in driven steady states remain elusive. We show that they are forbidden by a parity selection rule for rotation-driven linear nonequilibrium steady states. Whenever the relaxation and diffusion matrices commute, the snapshot mutual information between two time slices is exactly even under drive reversal, and parity violation rises linearly in the commutator norm when alignment is broken. Full isotropy strengthens this to drive-independence, and the planar mutual information takes the closed-form value of about 0.145 nats. Under the same alignment, the entropy production is exactly quadratic in the drive, and its prefactor admits an explicit closed form in the traces and determinant of the two matrices. The orthogonality of even and odd sectors leaves only one-sided thermodynamic-uncertainty bounds. The rule rests on the rotational symmetry of the drift alone and survives heavy-tailed isotropic stable noise with tail index below two, where variance-based bounds become vacuous. A falsifiable test is proposed on an electrical Brownian gyrator augmented for independent drive control with circuit-level stable-noise injection.

cs.IT↗

arXiv2026

FRACTAL: SSM with Fractional Recurrent Architecture for Computational Temporal Analysis of Long Sequences

Effective sequence modeling fundamentally requires balancing the retention of unbounded history with the high-resolution detection of abrupt short-term variations common in real-world phenomena. However, existing state space models (SSMs) relying on high-order polynomial projection operators (HiPPO) face a critical trade-off where uniform measures dilute recent information to maintain timescale invariance, while exponential measures sacrifice global context to capture local dynamics. This paper proposes a Fractional Recurrent Architecture for Computational Temporal Analysis of Long sequences (FRACTAL), a novel architecture integrating fractional measure theory into recursive memory updates to address this limitation. By deriving projection operators with analytically characterized spectral properties and a tunable singularity index, the proposed method amplifies sensitivity to recent signal perturbations while preserving the spectral structure that encodes scale-invariant memory dynamics. This theoretical innovation is instantiated within a simplified diagonalized state space framework by modulating input projection initialization to enable simultaneous capture of multi-scale temporal features. FRACTAL achieves an average score of 87.11\% on the Long Range Arena benchmark, including 61.85\% on the ListOps task, outperforming the S5 model.

cs.AI↗

arXiv2025

Beyond Gaussian Assumptions: A General Fractional HJB Control Framework for L\'evy-Driven Heavy-Tailed Channels in 6G

Emerging 6G wireless systems suffer severe performance degradation in challenging environments like high-speed trains traversing dense urban corridors and Unmanned Aerial Vehicles (UAVs) links over mountainous terrain. These scenarios exhibit non-Gaussian, non-stationary channels with heavy-tailed fading and abrupt signal fluctuations. To address these challenges, this paper proposes a novel wireless channel model based on symmetric $\alpha$-stable L\'evy processes, thereby enabling continuous-time state-space characterization of both long-term and short-term fading. Building on this model, a generalized optimal control framework is developed via a fractional Hamilton-Jacobi-Bellman (HJB) equation that incorporates the Riesz fractional operator to capture non-local spatial effects and memory-dependent dynamics. The existence and uniqueness of viscosity solutions to the fractional HJB equation are rigorously established, thus ensuring the theoretical validity of the proposed control formulation. Numerical simulations conducted in a multi-cell, multi-user downlink setting demonstrate the effectiveness of the fractional HJB-based strategy in optimizing transmission power under heavy-tailed co-channel and multi-user interference.

eess.SY↗

arXiv2025

A Model Can Help Itself: Reward-Free Self-Training for LLM Reasoning

Can language models improve their reasoning performance without external rewards, using only their own sampled responses for training? We show that they can. We propose Self-evolving Post-Training (SePT), a simple post-training method that alternates between self-generation and training on self-generated responses. It repeatedly samples questions, uses the model itself to generate responses under a specified sampling temperature, and then trains the model on the self-generated data. In this self-training loop, we use an online data refresh mechanism, where each new batch is generated by the most recently updated model. Across six math reasoning benchmarks, SePT improves a strong no-training baseline, defined as the untuned base model evaluated at its best swept decoding temperature, on several tested models. Additional ablations demonstrate the importance of online data refresh and temperature dynamics. Overall, our results identify a practical regime where reasoning can be improved using self-generated supervision alone. Our code is available at https://github.com/ElementQiii/SePT.

cs.LG↗

arXiv2025

Monolithic low-noise erbium-doped thin-film lithium niobate waveguide amplifier with 18 dB fiber to fiber net gain

Erbium-doped waveguide amplifiers have captured great attentions in recent years due to the rapid advance of photonic integration materials and fabrication techniques. In this work, a compact erbium-doped thin-film lithium niobate waveguide amplifier integrated with high-efficiency edge-couplers on the small footprint of 2 mm{\cross}25 mm, achieving >18 dB fiber-to-fiber (off-chip) net gain with bidirectional pumping by ~1480 nm laser diodes, is fabricated by the photolithography assisted chemo-mechanical etching technique. The fiber-to-fiber noise figures of the amplifier are also characterized to be around 5 dB, and the maximum amplified signal powers at the output fiber are above 13 dBm. Theoretical amplifier modeling resolving the erbium absorption and emission spectra predicts the efficient gain scaling with waveguide length for most of the telecom C-band wavelengths. The demonstrated high-external-gain erbium-doped waveguide amplifier will benefit various applications from optical communication and metrology to integrated photonic computing and artificial intelligence

physics.optics↗

arXiv2025

Fragility in Average Treatment Effect on the Treated under Limited Covariate Support

This paper studies the identification of the average treatment effect on the treated (ATT) under unconfoundedness when covariate overlap is partial. A formal diagnostic is proposed to characterize empirical support -- the subset of the covariate space where ATT is point-identified due to the presence of comparable untreated units. Where support is absent, standard estimators remain computable but cease to identify meaningful causal parameters. A general sensitivity framework is developed, indexing identified sets by curvature constraints on the selection mechanism. This yields a structural selection frontier tracing the trade-off between assumption strength and inferential precision. Two diagnostic statistics are introduced: the minimum assumption strength for sign identification (MAS-SI), and a fragility index that quantifies the minimal deviation from ignorability required to overturn qualitative conclusions. Applied to the LaLonde (1986) dataset, the framework reveals that nearly half the treated strata lack empirical support, rendering the ATT undefined in those regions. Simulations confirm that ATT estimates may be stable in magnitude yet fragile in epistemic content. These findings reframe overlap not as a regularity condition but as a prerequisite for identification, and recast sensitivity analysis as integral to empirical credibility rather than auxiliary robustness.

econ.EM↗

arXiv2025

StreamBP: Memory-Efficient Exact Backpropagation for Long Sequence Training of LLMs

Training language models on long sequence data is a demanding requirement for enhancing the model's capability on complex tasks, e.g., long-chain reasoning. However, as the sequence length scales up, the memory cost for storing activation values becomes huge during the Backpropagation (BP) process, even with the application of gradient checkpointing technique. To tackle this challenge, we propose a memory-efficient and exact BP method called StreamBP, which performs a linear decomposition of the chain rule along the sequence dimension in a layer-wise manner, significantly reducing the memory cost of activation values and logits. The proposed method is applicable to common objectives such as SFT, GRPO, and DPO. From an implementation perspective, StreamBP achieves less computational FLOPs and faster BP speed by leveraging the causal structure of the language model. Compared to gradient checkpointing, StreamBP scales up the maximum sequence length of BP by 2.8-5.5 times larger, while using comparable or even less BP time. Note that StreamBP's sequence length scaling ability can be directly transferred to batch size scaling for accelerating training. We further develop a communication-efficient distributed StreamBP to effectively support multi-GPU training and broaden its applicability. Our code can be easily integrated into the training pipeline of any transformer models and is available at https://github.com/Ledzy/StreamBP.

cs.LG↗

arXiv2024

GT23D-Bench: A Comprehensive General Text-to-3D Generation Benchmark

Text-to-3D (T23D) generation has emerged as a crucial visual generation task, aiming at synthesizing 3D content from textual descriptions. Studies of this task are currently shifting from per-scene T23D, which requires optimization of the model for every content generated, to General T23D (GT23D), which requires only one pre-trained model to generate different content without re-optimization, for more generalized and efficient 3D generation. Despite notable advancements, GT23D is severely bottlenecked by two interconnected challenges: the lack of high-quality, large-scale training data and the prevalence of evaluation metrics that overlook intrinsic 3D properties. Existing datasets often suffer from incomplete annotations, noisy organization, and inconsistent quality, while current evaluations rely heavily on 2D image-text similarity or scoring, failing to thoroughly assess 3D geometric integrity and semantic relevance. To address these fundamental gaps, we introduce GT23D-Bench, the first comprehensive benchmark specifically designed for GT23D training and evaluation. We first construct a high-quality dataset of 400K 3D assets, featuring diverse visual annotations (70M+ visual samples) and multi-granularity hierarchical captions (1M+ descriptions) to foster robust semantic learning. Second, we propose a comprehensive evaluation suite with 10 metrics assessing both text-3D alignment and 3D visual quality at multiple levels. Crucially, we demonstrate through rigorous experiments that our proposed metrics exhibit significantly higher correlation with human judgment compared to existing methods. Our in-depth analysis of eight leading GT23D models using this benchmark provides the community with critical insights into current model capabilities and their shared failure modes. GT23D-Bench will be publicly available to facilitate rigorous and reproducible research.

cs.CV↗

arXiv2024

On-chip wavelength division multiplexing by angled multimode interferometer fabricated on erbium-doped thin film lithium niobate on insulator

Photonic integrated circuits based on erbium doped thin film lithium niobate on insulator has attracted broad interests with insofar various waveguide amplifiers and microlasers demonstrated. Wideband operation facilitated by the broadband absorption and emission of erbium ions necessitates the functional integration of wavelength filter and multiplexer on the same chip. Here a low-loss wavelength division multiplexer at the resonant pumping and emission wavelengths (~1480 nm and 1530~1560 nm) of erbium ions based on angled multimode interferometer, is realized in the erbium doped thin film lithium niobate on insulator fabricated by the photolithography assisted chemomechanical etching technique. The minimum on-chip insertion losses of the fabricated device are <0.7 dB for both wavelength ranges, and a 3-dB bandwidth of >20 nm is measured at the telecom C-band. Besides, direct visualization of the multimode interference pattern by the visible upconversion fluorescence of erbium ions compares well with the simulated light propagation in the multimode interferometer. Spectral tuning of the wavelength division multiplexer by structural design is also demonstrated and discussed.

physics.optics↗

arXiv2022

Two-stage Contextual Transformer-based Convolutional Neural Network for Airway Extraction from CT Images

Accurate airway extraction from computed tomography (CT) images is a critical step for planning navigation bronchoscopy and quantitative assessment of airway-related chronic obstructive pulmonary disease (COPD). The existing methods are challenging to sufficiently segment the airway, especially the high-generation airway, with the constraint of the limited label and cannot meet the clinical use in COPD. We propose a novel two-stage 3D contextual transformer-based U-Net for airway segmentation using CT images. The method consists of two stages, performing initial and refined airway segmentation. The two-stage model shares the same subnetwork with different airway masks as input. Contextual transformer block is performed both in the encoder and decoder path of the subnetwork to finish high-quality airway segmentation effectively. In the first stage, the total airway mask and CT images are provided to the subnetwork, and the intrapulmonary airway mask and corresponding CT scans to the subnetwork in the second stage. Then the predictions of the two-stage method are merged as the final prediction. Extensive experiments were performed on in-house and multiple public datasets. Quantitative and qualitative analysis demonstrate that our proposed method extracted much more branches and lengths of the tree while accomplishing state-of-the-art airway segmentation performance. The code is available at https://github.com/zhaozsq/airway_segmentation.

eess.IV↗