arXiv subjects

Bo Gao

Publications and source records attributed to Bo Gao.

At least 19 recordsLinked to original sources

arXiv2026

End-to-End Stability of Online Active Learning in Large-Scale Neural Systems

Online Active Learning (OAL) aims to extend deep learning from static inference after offline training to continuous adaptation during real-time interaction with a changing physical environment. This transition is essential for safety-critical applications, but it raises a central theoretical question: whether the online parameter update process remains stable when scaled to multilayer neural architectures. This paper develops a stability-focused OAL framework. The framework uses prediction error as a feedback signal to update a structured deep model, thereby linking online learning with Lyapunov-based energy minimization. We show that prediction error convergence alone is insufficient for global parameter convergence because multilayer networks may admit high-dimensional manifolds of equivalent output solutions. The analysis therefore separates output error regulation, internal zero dynamics, interlayer coupling, and parameter convergence. Persistency of excitation and active damping are then used to remove the residual null-space modes that remain after output error convergence. A cascaded Lyapunov analysis establishes asymptotic stability of the overall OAL learning process, while monotonic activation functions and diffeomorphic layer mappings provide the uniqueness condition for the target parameter trajectory. The resulting proof provides a theoretical basis for applying online adaptive learning to large-scale neural systems in nonstationary and safety-critical settings.

cs.LG↗

arXiv2026

ReCast: Contract-Preserving Protection for Fixed-Interface Multimodal Reasoning

Remote multimodal models offer strong numerical reasoning capabilities over charts and speech, but sending private inputs risks exposing sensitive content. Text-only sanitization cannot directly satisfy fixed media interfaces, while identity anonymization leaves the underlying task content exposed. We introduce ReCast, an agentic plug-in framework that replaces source-specific content while preserving task-relevant relations and the required input modality. ReCast locally converts inputs into a shared textual evidence-query record, jointly rewrites entities and topics with a distilled 4B model, and substitutes values through a locally invertible, role-aware numerical map. A reconstruction agent generates and validates the required media from the protected record. The remote solver returns a program whose protected operands are restored locally before execution. On 4,000 held-out ChartQA and NMSQA examples, ReCast achieves 75.10% accuracy, retaining 92.43% of unprotected remote accuracy, while a model-based audit flags source-content leakage in 7.95% of solver-bound requests. It outperforms all evaluated local baselines, preserving the benefit of remote reasoning while reducing source-content exposure under existing media interfaces.

cs.CR↗

arXiv2026

Reward-Decomposed Reinforcement Learning for Immersive Video Role-Playing

Text-based role-playing models can imitate character styles, but often fail to capture scene atmosphere and evolving tension, which are crucial for immersive applications such as VR games and interactive narratives. We study video-grounded role-playing dialogue and introduce EBM-RL (Eye--Brain--Mouth Reinforcement Learning), a decoupled GRPO-based framework that separates observation ( ), reasoning ( ), and utterance generation ( ). This design mimics the human See-Think-Speak process, enabling the model to ground dialogue in visual perception before reasoning and response generation. To optimize this See-Think-Speak process, EBM-RL integrates complementary rewards for scene--text alignment, perceptual--cognitive utility, answer faithfulness, and format consistency. Extensive experiments show that EBM-RL substantially outperforms text-only role-playing baselines and larger-scale vision-language models on our immersive role-playing benchmark, improving both visual-atmosphere consistency and character authenticity. Moreover, EBM-RL demonstrates strong zero-shot transfer to out-of-domain VideoQA benchmarks without additional fine-tuning. We also release an open-source dataset for video-grounded role-playing dialogue.

cs.AI↗

arXiv2026

Updated estimates of post-merger gravitational wave energy released in GW170817 and the maximum neutron star mass

In this work, benefiting from the increased sample of neutron stars with measured masses and radii as well as the incorporation of the chiral effective field theory and perturbative QCD constraints, the tidal parameter $κ_2^T$ is constrained to be $78^{+17}_{-11}$ (68.3% credible interval, mainly adopted in this work unless mentioned specifically) for the binary neutron stars involved in GW170817. Such a $κ_2^T$ is in favor of strong gravitational wave radiation in the post-merger phase and the corresponding energy is estimated to be $E_{\rm GW,p} \simeq 0.051^{+0.022}_{-0.017}\,M_\odot c^2$. Assuming the remnant from binary neutron star merger is a supramassive neutron star, as suggested by the modeling of the electromagnetic counterparts of GW170817, we examine the maximum mass of nonrotating neutron stars ($M_{\rm TOV}$) while accounting for the uncertainty in the remnant's lifetime ($t_{\rm c}$). Our results show that $M_{\rm TOV}$ varies from $2.09^{+0.11}_{-0.09}\,M_\odot$ for $t_{\rm c}=0.1$ s to $2.18^{+0.10}_{-0.09}\,M_\odot$ for $t_{\rm c}=1$ s. This result is consistent with that independently inferred from the re-construction of the equation of state of neutron star matter, i.e., $M_{\rm TOV,exc}=2.16^{+0.10}_{-0.07}M_\odot$, particularly if the very massive neutron stars with masses measured indirectly have been removed in constructing the prior distribution of $M_{\rm TOV}$. The consistency of the maximum mass of nonrotating neutron stars found in different approaches suggests a reasonable understanding of this key parameter for dense-matter physics.

astro-ph.HE↗

arXiv2026

ReGraP-LLaVA: Reasoning enabled Graph-based Personalized Large Language and Vision Assistant

Multimodal Large Language Models have shown strong performance across multimodal tasks, and recent personalized MLLMs can recognize user-specific concepts and generate contextual captions. However, existing personalized MLLMs mainly focus on isolated concepts, often lacking relational training data, neglecting connections among personalized concepts, and evaluating mostly on recognition or captioning. To address these limitations, we introduce ReGraP, a dataset of 120 personalized knowledge sets, each containing images, knowledge graphs, and Chain-of-Thought Question-Answering pairs. Based on ReGraP, we propose Reasoning enabled Graph-based Personalized Large Language and Vision Assistant ReGraP-LLaVA, a personalized MLLM that incorporates KGs and CoT QA pairs through soft and/or hard graph prompting to align structured relational knowledge with the model's semantic space. We further establish the ReGraP Benchmark, covering multiple-choice, fill-in-the-blank, true/false, and descriptive questions in both open- and closed-ended settings, to evaluate personalized relational reasoning and knowledge-connection capabilities. Experimental results show that ReGraP-LLaVA effectively learns personalized knowledge and performs relational reasoning, achieving the best overall performance among competitive baselines. Code and data are available at: https://github.com/xyfyyds/ReGraP

cs.CV↗

arXiv2026

Improving O-RADS Risk Stratification from Ultrasound Reports: A Comparative Evaluation of Hybrid versus End-to-End LLM Reasoning Strategies

Background: Automating clinical guideline-based decision-making with large language models (LLMs) remains challenging because of reliability, hallucination, and limited interpretability. We compared the performance of LLMs and reasoning strategies for automated Ovarian-Adnexal Reporting and Data System (O-RADS) classification from free-text pelvic ultrasound reports. Methods: In this retrospective study, consecutive patients with ovarian masses who underwent pelvic ultrasound were included. Eight LLMs were tested with three reasoning strategies: implicit-knowledge end-to-end, rule-informed end-to-end, and a feature-based hybrid architecture that decoupled feature extraction from rule-based classification. The reference standard was O-RADS categorization established by expert consensus. Results: A total of 310 women with 390 ovarian masses were evaluated. The feature-based hybrid architecture using Gemini 3.6 Flash demonstrated the best performance, achieving an accuracy of 99.2% (387 of 390) and almost perfect agreement with the reference standard (weighted kappa = 1.00; 95% CI: 0.99-1.00). Its performance surpassed that of original clinical reports (accuracy, 87.7% [342 of 390]; weighted kappa = 0.94; 95% CI: 0.91-0.96) and end-to-end LLM strategies (accuracy range, 65.6% [256 of 390] to 95.9% [374 of 390]). For structured feature extraction, Gemini 3.6 Flash demonstrated higher overall accuracy than Claude Fable 5 (98.9% vs 97.8%; P < 0.001). The hybrid architecture reduced misclassification errors and mitigated the overstaging tendency observed in original reports. Conclusion: The feature-based hybrid LLM architecture that separates clinical feature extraction from deterministic guideline execution enables highly accurate, reliable, and interpretable automated O-RADS classification, providing a promising approach for standardized, guideline-based clinical decision-making.

cs.AI↗

arXiv2026

Enhanced Rydberg-Atom Superheterodyne Detection of Hidden-Photon Dark Matter on Chips

Although hidden-photon dark matter with masses above $10^{-4}\,\mathrm{eV}$ is well motivated by inflationary production, it remains largely unexplored by terrestrial experiments. Through kinetic mixing, hidden photons induce a weak oscillating electric field above $10\,\mathrm{GHz}$. We propose to amplify this signal using a compact high-frequency distributed cavity and detect it with chip-scale Rydberg-atom superheterodyne spectroscopy. Combining resonant enhancement, large dipole moments of Rydberg atoms, and long-term stable integration, this approach can probe hidden-photon dark matter in the mass range $5 \times 10^{-5}\text{--}7\times 10^{-4}\,\mathrm{eV}$ with sensitivities $3$--$4$ orders of magnitude beyond existing limits.

hep-ph↗

arXiv2026

UniMamba: A Unified Spatial-Temporal Modeling Framework with State-Space and Attention Integration

Multivariate time series forecasting is fundamental to numerous domains such as energy, finance, and environmental monitoring, where complex temporal dependencies and cross-variable interactions pose enduring challenges. Existing Transformer-based methods capture temporal correlations through attention mechanisms but suffer from quadratic computational cost, while state-space models like Mamba achieve efficient long-context modeling yet lack explicit temporal pattern recognition. Therefore we introduce UniMamba, a unified spatial-temporal forecasting framework that integrates efficient state-space dynamics with attention-based dependency learning. UniMamba employs a Mamba Variate-Channel Encoding Layer enhanced with FFT-Laplace Transform and TCN to capture global temporal dependencies, and a Spatial Temporal Attention Layer to jointly model inter-variate correlations and temporal evolution. A Feedforward Temporal Dynamics Layer further fuses continuous and discrete contexts for accurate forecasting. Comprehensive experiments on eight public benchmark datasets demonstrate that UniMamba consistently outperforms state-of-the-art forecasting models in both forecasting accuracy and computational efficiency, establishing a scalable and robust solution for long-sequence multivariate time-series prediction.

cs.LG↗

arXiv2026

Steering topology distributions for unified generative design of architected metamaterials

Architected metamaterials derive their functions from structure, creating vast opportunities to program physical responses through topology design. However, existing design methods are often tailored to individual design problems, making limited use of topology knowledge for effective and broadly applicable design as objectives, constraints, and physical functions change. Here we introduce Generative Topology Optimization (GenTO), a unified framework that turns a learned topology prior into a reusable design engine. GenTO trains a diffusion model on a large full-order topology dataset and then iteratively steers the resulting topology distribution toward task-specific high-performing regions using user-defined physical objectives and constraints. This shifts the object of optimization from a single structure to a task-adapted topology distribution. Across topology design problems spanning thermal extremization, multi-objective morphology control, property-targeted auxetic design, and vibration transmission design, GenTO reuses pretrained topology priors for heterogeneous tasks, preserves structural diversity, and reaches high-performing solutions supported by numerical benchmarks and experimental validation. These results establish reusable topology knowledge as a unified principle for effective and scalable architected metamaterial design.

cs.AI↗

arXiv2026

Softplus Attention with Re-weighting Boosts Length Extrapolation in Large Language Models

Large language models have achieved remarkable success in recent years, primarily due to self-attention. However, traditional Softmax attention suffers from numerical instability and reduced performance as the number of inference tokens increases. This work addresses these issues by proposing a new design principle for attention, viewing it as a two-stage process. The first stage (normalisation) refines standard attention by replacing Softmax with the more numerically stable Softplus followed by $l_{1}$-normalisation. Furthermore, we introduce a dynamic scale factor based on invariance entropy. We show that this novel attention mechanism outperforms conventional Softmax attention, and state-of-the-art Softmax-free alternatives. Our second proposal is to introduce a second processing stage (sharpening) which consists of a re-weighting mechanism that amplifies significant attentional weights while diminishing weaker ones. This enables the model to concentrate more effectively on relevant tokens, mitigating the attention sink phenomenon, and fundamentally improving length extrapolation. This novel, two-stage, replacement for self-attention is shown to ensure numerical stability and dramatically improve length extrapolation, maintaining a nearly constant validation loss at 16$\times$ the training length while achieving superior results on challenging long-context retrieval tasks and downstream benchmarks. Furthermore, symbolic regression experiments demonstrate that our method enables models to recover Newton's gravitational law from orbital trajectory sequences, providing evidence that appropriate attention mechanisms are crucial for foundation models to develop genuine physical world models. Our code is available at https://github.com/iminfine/freeattn.

cs.CL↗

arXiv2026

A Symmetric Superconducting Dome Hosts Non-Fermi Liquid Behavior at Optimal Doping in MoS2

The similarities between the phase diagrams of ionic liquid-gated transition metal dichalcogenides (TMDCs) and high-temperature superconductors have garnered considerable interest due to the presence of a superconducting dome with a non-monotonic dependence of the superconducting order parameter as a function of charge carrier density. However, the lack of a complete superconducting dome and insights into the normal state in ionic liquid-gated TMDCs prevents a detailed mapping between the two systems. In this work, we obtain a symmetric superconducting dome that extends from deep underdoped regime all the way to deep overdoped regime in ionic liquid gated MoS2 with a refined gating protocol. We demonstrate that the dome is anticorrelated with the evolution of non-Fermi liquid behavior in the normal states. The scattering rate in the non-Fermi liquid regime can reach Planckian limit. The results may shed light on the emergence of superconductivity in TMDCs.

cond-mat.supr-con↗

arXiv2026

DFIR-DETR: Frequency-Domain Iterative Refinement and Dynamic Feature Aggregation for Small Object Detection

Small object detection in complex scenes exposes a fundamental tension in neural network design: backbone attention distributes computation uniformly regardless of content, pyramid necks inflate activation magnitudes during upsampling without norm compensation, and bottleneck convolutions progressively smooth high-frequency edge components through accumulated spatial filtering. In response, we develop DFIR-DETR by tracing each proposed module back to a specific, measurable deficiency in the RT-DETR baseline: uniform attention that ignores spatial complexity, norm drift that destabilises upsampled features, and spatial convolutions that progressively suppress the high-frequency components small objects depend on. On NEU-DET and VisDrone, DFIR-DETR achieves 92.9% and 51.6% mAP50 with only 11.7M parameters and 47.2 GFLOPs, demonstrating consistent gains across two qualitatively different detection domains.

cs.CV↗

arXiv2026

PCDM: A Diffusion-Based Data Poisoning Attack Against Federated Learning Systems

Federated learning (FL) is vulnerable to data poisoning attacks due to its distributed nature. Although recent GAN-based data poisoning methods have indicated the potential of using generative AI to generate seemingly legitimate poisoned data, the inherent consistency of GAN outputs can still reveal a sign of data poisoning. In this paper, we propose a diffusion-based data poisoning framework against FL systems, which leverages a Poisoning-Oriented Conditional Diffusion Model (PCDM) to enable fine-grained control over the local generation of poisoned data while ensuring both attack effectiveness and stealthiness. Our PCDM incorporates an adjustable poisoning vector within the global context to precisely control the generation of poisoned data, with theoretical guarantees on attack performance. Furthermore, it employs a novel jumping diffusion strategy for lightweight and efficient poisoned data generation. We conduct the most systematic and broad experimental evaluation for FL poisoning attacks against various defenses, including advanced Byzantine robust aggregation mechanisms, on four open datasets: MNIST, Fashion-MNIST, CIFAR-10, CIFAR-100, and a real-world wireless-specific dataset VRAI. Our results demonstrate that PCDM is less likely to exhibit statistical anomalies compared with the state-of-the-art methods while more effectively degrading global FL performance, which poses a significant risk to data security in FL.

cs.CR↗

arXiv2026

Testing Electromagnetic Memory via Acceleration-Induced Phase Imprints in Superconductors

Electromagnetic memory is an infrared observable of gauge theory associated with soft photons and large gauge transformations. Despite its fundamental theoretical importance, it has not yet been experimentally verified. From a phenomenological perspective, a transient electromagnetic configuration can leave a persistent gauge-invariant phase imprint on charged coherent states after the local field has vanished. We point out that the electric field and associated gauge potential induced inside a normal conductor by gravitational acceleration can provide a clean source for imprinting this phase, and it can then be read out through a superconducting protocol. For representative parameters, the predicted signal can lie within the range of present sensitivities, providing a possible tabletop route toward testing electromagnetic memory.

hep-ph↗

arXiv2026

BOOKAGENT: Orchestrating Safety-Aware Visual Narratives via Multi-Agent Cognitive Calibration

Recent advancements in Large Generative Models (LGMs) have revolutionized multi-modal generation. However, generating illustrated storybooks remains an open challenge, where prior works mainly decompose this task into separate stages, and thus, holistic multi-modal grounding remains limited. Besides, while safety alignment is studied for text- or image-only generation, existing works rarely integrate child-specific safety constraints into narrative planning and sequence-level multi-modal verification. To address these limitations, we propose BookAgent, a safety-aware multi-agent collaboration framework designed for high-quality, safety-aware visual narratives. Different from prior story visualization models that assume a fixed storyline sequence, BookAgent targets end-to-end storybook synthesis from a user draft by jointly planning, scripting, illustrating, and globally repairing inconsistencies. To ensure precise multi-modal grounding, BookAgent dynamically calibrates page-level alignment between textual scripts and visual layouts. Furthermore, BookAgent calibrates holistic consistency from the temporal dimension, by verifying-then-rectifying global inconsistencies in character identity and storytelling logic. Extensive experiments demonstrate that BookAgent significantly outperforms current methods in narrative coherence, visual consistency, and safety compliance, offering a robust paradigm for reliable agents in complex multi-modal creation. The implementation will be publicly released at https://github.com/bogao-code/BookAgent/tree/main.

cs.CV↗

arXiv2026

CoT2-Meta: Budgeted Metacognitive Control for Test-Time Reasoning

Recent test-time reasoning methods improve performance by generating more candidate chains or searching over larger reasoning trees, but they typically lack explicit control over when to expand, what to prune, how to repair, and when to abstain. We introduce CoT2-Meta, a training-free metacognitive reasoning framework that combines object-level chain-of-thought generation with meta-level control over partial reasoning trajectories. The framework integrates four components: strategy-conditioned thought generation, tree-structured search, an online process oracle for step-level reasoning evaluation, and a meta-controller that allocates computation through expansion, pruning, repair, stopping, and fallback decisions. Under matched inference budgets, CoT2-Meta consistently outperforms strong single-path, sampling-based, and search-based baselines, including ReST-MCTS. On the default backbone, it achieves 92.8 EM on MATH, 90.4 accuracy on GPQA, 98.65 EM on GSM8K, 75.8 accuracy on BBEH, 85.6 accuracy on MMMU-Pro, and 48.8 accuracy on HLE, with gains over the strongest non-CoT2-Meta baseline of +3.6, +5.2, +1.15, +2.0, +4.3, and +4.3 points, respectively. Beyond these core results, the framework remains effective across a broader 15-benchmark suite spanning knowledge and QA, multi-hop reasoning, coding, and out-of-distribution evaluation. Additional analyses show better compute scaling, improved calibration, stronger selective prediction, targeted repair behavior, and consistent gains across backbone families. These results suggest that explicit metacognitive control is a practical design principle for reliable and compute-efficient test-time reasoning systems.

cs.AI↗

arXiv2026

A Model Consistency-Based Countermeasure to GAN-Based Data Poisoning Attack in Federated Learning

In federated learning (FL), although the original intention of available but not visible data is to allay data privacy concerns, it potentially brings new security threats, particularly poisoning attacks that target such not visible local data. Intuitively, such data poisoning attacks have great potential in stealthily degrading global FL outcomes, and are expected to be even stealthier if being enhanced by generative models like generative adversarial networks (GANs). However, existing defense methods have not been thoroughly challenged in this regard and generally fail to be aware of a local generation of seemingly legitimate poisoned data. With a growing concern on potentially stealthier attacks, in this paper, a cost-effective defense mechanism named Model Consistency-Based Defense (MCD) is proposed, which offers a comprehensive examination of available local models across multiple feature dimensions, providing an indirect yet effective means of identifying hidden data poisoning attackers. To push the limit of MCD against stealthier attacks, we propose a new GAN-based data poisoning attack model named VagueGAN and an unsupervised variant of it, which can be flexibly deployed to generate seemingly legitimate but noisy poisoned data. The consistency of GAN outputs revealed by VagueGAN helps strengthen MCD to work against stealthier GAN-based attacks as well as other mainstream ones. Extensive experiments on multiple open datasets (MNIST, Fashion-MNIST, CIFAR-10, CIFAR-100, and Mini-Imagenet) indicate that our attack method better balances the trade-off between attack effectiveness and stealthiness with low complexity. More importantly, our defense mechanism is shown to be more competent in identifying a variety of poisoned data, particularly stealthier GAN-poisoned ones.

cs.CR↗

arXiv2026

All-optical intracellular thermal profiling using nanodiamond-based "thermal radar"

The local thermal conductivity (\k{appa}) is a pivotal biophysical parameter, governing intracellular heat flux and underlying functional processes like metabolic regulation and stress response. However, label-free mapping with sub-micron resolution in living cells remains challenge. Here, we present frequency-domain fluorescence thermometry (FD-FTM), an all-optical method based on a hybrid nanodiamond-on-gold-membrane platform, which enables quantitative mapping of \k{appa} in biological systems. Fluorescence nanodiamonds (FNDs) are deposited on substrates coated with a 50 nm gold membrane, where FNDs function as nanoscale thermometers, and the gold membrane serves as a photothermal heat source. We validate FD-FTM across reference materials and biological media, with fitting uncertainties of ~10%. By varying the modulation frequency, we tune the thermal penetration depths, enabling controlled heat propagation from the substrate to the cell nucleus. The method delivers sensitivity sufficient to resolve changes in biofluid thermal conductivity on the order of 16% relative to water. Using these capabilities, we demonstrate non-invasive thermal profiling across scales: at the cellular level, nuclear chromatin packing yields \k{appa} higher by ~10% relative to the cytoplasm; at the organelle level, we resolve \k{appa} variations associated with protein aggregates formed during liquid-liquid phase separation in an amyotrophic lateral sclerosis disease model. Temporal measurements in living cells over 30 minutes further reveal spatially resolved intracellular responses to osmotic stress, linking nanoscale thermal dynamics to biomolecular condensates. These results establish FD-FTM as a label-free, robust, and quantitative platform for thermally decoding intracellular processes, opening avenues for studying metabolic heterogeneity, disease mechanisms, and therapeutic responses.

physics.bio-ph↗