Searcharxiv⌕ Search

arXiv subjects

Yaochu Jin

Publications and source records attributed to Yaochu Jin.

At least 19 recordsLinked to original sources

Secure Multimodal Federated Optimization via Hierarchical DNA-Based Key Management

Federated optimization over heterogeneous multimodal data enables distributed institutions to collaboratively optimize a shared model using medical images, genomic profiles, clinical records, and physiological signals without centralizing raw data. However, protecting modality-specific data throughout the optimization lifecycle remains insufficiently explored, particularly when heterogeneous modalities require separate encryption keys and selective local access. This paper proposes a secure multimodal federated optimization framework that separates local data protection from federated-update protection. A trusted key-management authority securely provisions one client key through Hybrid Public Key Encryption (HPKE), from which each client locally derives modality-specific symmetric keys using the domain-separated HMAC-based Extract-and-Expand Key Derivation Function (HKDF). Heterogeneous modalities are mapped into a unified DNA-domain representation and processed through a DNA-origami-inspired structural transformation before authenticated encryption with AES-GCM. During local optimization, only the required modalities are decrypted within an authorized client environment. Individual model updates are independently protected through secure aggregation, allowing the coordinator to observe only aggregated updates rather than each client's contribution. Together, these components establish a structured approach to modality-specific protection and secure collaborative optimization, integrating hierarchical key management, authenticated local encryption, and aggregation-level update protection within a unified framework.

cs.CR↗

Beyond Fixed Visual History: Learning Motion-Adaptive Temporal Receptive Fields for Complex-Terrain Quadruped Locomotion

Versatile quadruped locomotion depends on a tight coupling between what the robot perceives and what its body can execute. Yet existing visual controllers typically organize depth history using fixed temporal patterns, overlooking that the same temporal span corresponds to different spatial ranges as the robot's speed, posture, and gait phase change. Consequently, the same visual history may be insufficient in one motion state but outdated or misaligned in another. To address this limitation, we propose TempoLoco, a body-aware dynamic visual perception framework that organizes visual memory according to the robot's current locomotion state. TempoLoco learns multiple complementary temporal views from a long depth buffer, encodes them within a shared feature space, and dynamically aggregates them using proprioceptive history and multi-scale terrain evidence. This enables a single policy to align visual anticipation with the robot's current motion capabilities across complex terrains, including stairs, platforms, gaps, discrete obstacles, rough ground, and slopes. From the experiment analysis, TempoLoco achieves a 78.4% success rate, compared with 58.2% for the strongest fixed-history baseline and 62.1% for uniform multi-scale fusion. In separate capability evaluations, it completes 41/45 simulated trials across nine terrain types and 23/25 real-world trials on a Unitree B2 with a fixed arm, demonstrating robust and deployable locomotion through body-aware dynamic visual perception.

cs.RO↗

Decoding One Safety Trigger Token for Balancing Safety and Usability in Large Language Models

Large Language Models (LLMs) have been extensively used across diverse domains, including virtual assistants, automated code generation, and scientific research. However, they remain vulnerable to jailbreak attacks, which manipulate the models into generating harmful responses despite safety alignment. Recent studies have shown that current safety-aligned LLMs undergo shallow safety alignment. In this work, we conduct an in-depth investigation into the underlying mechanism of this phenomenon and reveal that it manifests through learned ''safety trigger tokens'' that activate the model's safety patterns when paired with the specific input. Through both analysis and empirical verification, we further demonstrate the high similarity of the safety trigger tokens across different harmful inputs. Accordingly, we propose D-STT, a simple yet effective defense algorithm that identifies and explicitly decodes safety trigger tokens of the given safety-aligned LLM to activate the model's learned safety patterns. In this process, the safety trigger is constrained to a single token, which effectively preserves model usability by introducing minimum intervention in the decoding process. Extensive experiments across diverse jailbreak attacks and benign prompts demonstrate that D-STT significantly reduces output harmfulness while preserving model usability and incurring negligible response time overhead, outperforming ten baseline methods.

cs.CR↗

DexWeave: Learning Dexterous Humanoid Loco-Manipulation from Human Demonstrations

Learning dexterous humanoid loco-manipulation from human demonstrations requires transferring not only human motion, but also the coordinated interaction structure underlying the demonstrated behavior. This is challenging because embodiment differences distort the coupling among body motion, wrist placement, finger articulation, and object interaction, while kinematically accurate references may still be difficult to realize under robot dynamics. We present DexWeave, a unified framework that connects interaction-consistent motion retargeting with anatomy-aware whole-body policy learning. DexWeave first employs a two-stage retargeting procedure that initializes body and hand motions with specialized solvers and subsequently performs coupled refinement over the upper-body interaction chain while preserving lower-body support. The resulting references are tracked by an anatomy-aware Transformer policy that represents anatomical regions as structured tokens and uses directed masked attention to model their dependencies, with object information selectively conditioning the upper-body pathway for dexterous interaction. The policy jointly outputs body and dexterous-hand actions and is trained directly with reinforcement learning, without pretrained tracking policies, teacher-student distillation, or subsequent residual refinement. DexWeave improves retargeting fidelity and interaction consistency while achieving higher manipulation performance and faster policy convergence than MLP baselines. We further deploy the learned policies on a physical Unitree G1 humanoid equipped with Inspire dexterous hands, demonstrating dexterous whole-body loco-manipulation in the real world. See our project page (https://dexweave.github.io) for videos.

cs.RO↗

SpikeCredit: Temporal Credit Carrier for Reinforcement Learning with Sparse Rewards

Reinforcement learning (RL) with sparse rewards is challenging because delayed outcomes provide little guidance about which intermediate computations caused success or failure. We argue that reliable credit assignment requires policy dynamics that preserve and expose credit-relevant information over time, a role we formalize as Temporal Credit Carriers (TCCs) and that spiking neural networks (SNNs) naturally fulfill through graded membrane traces and event-driven spikes. Based on this hypothesis, we propose SpikeCredit, an SNN-based framework for RL with sparse rewards that first performs task-adaptive TCC selection and then closes the loop between a fast TCC-reading pathway, where self-motion feedback constraint uses local behavior-grounded cues to constrain transition-level credit recovery, and a slow TCC-writing pathway, where credit-targeted trace alignment feeds recovered credit back into the actor to make future TCC dynamics more credit-readable. Across sparse-reward MuJoCo tasks, SpikeCredit improves Last10 return over sparse SNN baselines by +1169% on Ant, +953% on Hopper, +723% on Swimmer, and +1781% on Walker2d, and exceeds the dense-reward baseline on Swimmer by +113%. Mechanistic analyses further show substantially stronger alignment with dense rewards than the sparse SNN baseline. These results position spiking dynamics as credit-preserving substrates for sparse-reward RL.

cs.LG↗

In-Context Learning for Robots: Methods and Applications

General-purpose robots must infer what a new task requires and translate that understanding into appropriate physical action. In-context learning (ICL) for robots supports this process by using demonstrations and interaction to direct existing competence with neural parameters held fixed during deployment. We organize this literature review around the interfaces connecting contextual evidence to execution, distinguishing four families: context-conditioned policies, geometric demonstration transfer, world-model-based control, and skill- and agent-based execution. Comparing these interfaces clarifies their transfer assumptions and the roles of training, correspondence, and memory in making context useful. Across manipulation and navigation, we examine how these mechanisms preserve taught requirements as objects, environments, and execution conditions change. This analysis links method design to evaluation practices that distinguish responsiveness to teaching, physical transfer, and benefits from retained experience. The resulting agenda connects compositional task acquisition and faithful transfer with physical recursive self-improvement, in which experience improves the ability to learn subsequent tasks.

cs.RO↗

IP Protection in the Era of Visual Generative AI: A Survey

The rapid evolution of visual generative AI has introduced a wide range of intellectual property risks, spanning the unauthorized learning, reproduction, extraction, misuse, and redistribution of protected data and model assets. To address these risks, a growing body of technical defenses has been proposed. However, existing surveys typically organize this literature by lifecycle stage or technical mechanism, which can obscure the protective intent of different methods. This survey presents a two-dimensional taxonomy for IP protection in visual generative models. The primary axis is a Control Logic View, which classifies methods into Information Exposure Control, Generative Behavior Constraint, and Attribution & Accountability according to the risk variable they regulate. The secondary axis distinguishes Data IP from Model IP as cross-cutting asset dimensions. Under this framework, we systematically review protection methods, align evaluation protocols with protection objectives, and discuss open challenges including proactive model-level safeguards, standardized evaluation, robustness against adaptive attacks, and explainable evidence. This survey aims to offer a principled, systematic, and easy-to-follow overview for both new and experienced researchers in visual generative AI IP protection.

cs.CV↗

Rethinking Attention Locality in Spiking Transformers

Spiking Transformers provide a promising paradigm for efficient visual processing with spike-driven computation, yet their Softmax-free Spiking Self-Attention (SSA) struggles to establish spatially localized token interactions. Although existing locality-enhanced SSA methods improve accuracy, it remains unclear whether they consistently induce spatial locality across layers and different Spiking Transformer architectures. Through Mean Attention Distance (MAD) analysis, we reveal that computational locality does not necessarily translate into spatial locality and show that uniformly applying the same locality enhancement overlooks architecture-dependent deployment requirements. Motivated by these observations, we propose Spatially Contiguous Local Attention with Boundary Continuity Pathway (SCLA-BCP). SCLA computes attention within non-overlapping regions of spatially adjacent tokens, while BCP facilitates cross-boundary information exchange through a lightweight convolutional pathway. Furthermore, we develop a hierarchical locality deployment strategy to effectively apply SCLA-BCP across the two major Spiking Transformer architectures. Extensive experiments on seven static and neuromorphic datasets covering classification, detection, and segmentation demonstrate consistent improvements with limited parameter and energy overhead. Notably, our approach improves mAP@50 by up to 9.50% on COCO 2017 and mIoU by up to 3.42% on ADE20K. Visualizations, MAD analysis, and ablation studies further validate its effectiveness.

cs.CV↗

STRONG-VLA: Decoupled Robustness Learning for Vision-Language-Action Models under Multimodal Perturbations

Despite their strong performance in embodied tasks, recent Vision-Language-Action (VLA) models remain highly fragile under multimodal perturbations, where visual corruption and linguistic noise jointly induce distribution shifts that degrade task-level execution. Existing robustness approaches typically rely on joint training with perturbed data, treating robustness as a static objective, which leads to conflicting optimization between robustness and task fidelity. In this work, we propose STRONG-VLA, a decoupled fine-tuning framework that explicitly separates robustness acquisition from task-aligned refinement. In Stage I, the model is exposed to a curriculum of multimodal perturbations with increasing difficulty, enabling progressive robustness learning under controlled distribution shifts. In Stage II, the model is re-aligned with clean task distributions to recover execution fidelity while preserving robustness. We further establish a comprehensive benchmark with 28 perturbation types spanning both textual and visual modalities, grounded in realistic sources of sensor noise, occlusion, and instruction corruption. Extensive experiments on the LIBERO benchmark show that STRONG-VLA consistently improves task success rates across multiple VLA architectures. On OpenVLA, our method achieves gains of up to 12.60% under seen perturbations and 7.77% under unseen perturbations. Notably, similar or larger improvements are observed on OpenVLA-OFT (+14.48% / +13.81%) and pi0 (+16.49% / +5.58%), demonstrating strong cross-architecture generalization. Real-world experiments on an AIRBOT robotic platform further validate its practical effectiveness. These results highlight the importance of decoupled optimization for multimodal robustness and establish STRONG-VLA as a simple yet principled framework for robust embodied control.

cs.RO↗

An Evolutionary Algorithm Assisted by an Ensemble of Pareto-Optimal Surrogate Models

An ensemble of surrogate models helps improve the prediction quality and robustness of surrogate models, and in turn, the search performance of surrogate-assisted evolutionary algorithms (SAEAs). Although different degrees of smoothness of the approximated fitness landscapes need to be carefully designed for an effective ensemble, little attention has been paid to the explicit tuning of the degree of smoothness derived by surrogate models. This study proposes an adaptive ensemble SAEA, which automatically constructs plausible ensemble models by optimizing their parameter settings. Unlike existing adaptive/ensemble SAEAs, which consider prediction accuracy alone, the proposed algorithm optimizes the structure of radial basis function networks (RBFNs) by solving bi-objective minimization problems of approximation error and model complexity, resulting in robust ensemble models of accurate surrogate models with different degrees of smoothness of the approximated fitness landscapes. As a result, the over/under-fittings are reduced. Additionally, an infill criterion is designed so that surrogate models with different degrees of smoothness can contribute to the solution prescreening. The experimental results demonstrated the statistical superiority of our algorithm over state-of-the-art SAEAs on a single-objective benchmark and real-world problem sets under an expensive optimization scenario. The source code of the proposed algorithm is available at https://github.com/haranychan/EPOS

cs.NE↗

Position Paper: Unlocking the Potential of AI Researchers in Scientific Discovery-What Is Missing?

The potential of AI researchers in scientific discovery remains largely untapped. Over the past decade, AI for Science (AI4Science) publications in 145 Nature Index journals have increased fifteen-fold, yet they still account for less than 3% of the total publications. Drawing upon the Diffusion of Innovation theory, we project AI4Science's share of total publications to rise from 2.72% in 2024 to approximately 20% by 2050. Achieving this shift requires fully harnessing the potential of AI researchers, as nearly 95% of AI-driven research in these journals is led by experimental scientists. To facilitate this, we propose structured workflows and strategic interventions to position AI researchers at the forefront of scientific discovery. Specifically, we identify three critical pathways: equipping experimental scientists with accessible AI tools to amplify the impact of AI researchers, bridging cognitive and methodological gaps to enable more direct involvement in scientific discovery, and proactively fostering a thriving AI-driven scientific ecosystem. By addressing these challenges, we aim to empower AI researchers as key drivers of future scientific breakthroughs.

cs.CY↗

Beyond Speedups: Hardware-Aware Evaluation of Evolutionary Algorithms on GPUs

Evolutionary algorithms (EAs) are increasingly executed on graphics processing units (GPUs) to exploit population-level parallelism. This shift changes the resource model under which EAs are designed and evaluated. However, many GPU-based EA studies still focus mainly on implementation-level speedup after porting CPU-oriented algorithms to GPUs, providing limited insight into how algorithmic mechanisms, function-evaluation (FE) budgets, population scales, and hardware utilization jointly affect optimization behavior. In response, this paper goes beyond speedup measurement and studies the scaling behavior of EAs on GPUs from a hardware-aware evaluation perspective. We evaluate 16 representative EAs on 30 benchmark problems across CPU and GPU platforms, covering single-objective optimization, multi-objective optimization, numerical benchmarks, and neuroevolution tasks. The study leads to four findings. First, GPU acceleration is highly heterogeneous across algorithms because different evolutionary mechanisms expose different degrees of batched computation, memory regularity, and synchronization. Second, FE-budgeted evaluation remains useful for measuring sample efficiency, but it provides only a limited observation window under GPU execution; time-budgeted evaluation is therefore necessary for assessing practical time-to-solution and long-horizon search behavior. Third, GPU effectiveness depends on scaling regimes induced by problem dimension and population size, where parallelism may be underutilized, effective, or saturated. Fourth, GPU execution makes very large populations practically affordable, and several evolutionary mechanisms can convert this increased population scale into improved optimization performance. These results indicate that GPU parallelism should not be treated only as a post hoc acceleration tool, but as part of the evaluation and design assumptions of scalable EAs.

cs.NE↗

Pruning Federated Models through Loss Landscape Analysis and Client Agreement Scoring

The practical deployment of Federated Learning (FL) on resource-constrained devices is fundamentally limited by the high cost of training large models and the instability caused by heterogeneous (non-IID) client data. Conventional pruning methods often treat data heterogeneity as a problem to be mitigated. In this work, we introduce a paradigm shift: we reframe client diversity as a feature to be harnessed. We propose AutoFLIP, a framework that begins not with training, but with a one-time federated loss exploration. During this phase, clients collaboratively build a map of the collective loss landscape, using their diverse data to reveal the problem's essential structure. This shared intelligence then guides an adaptive pruning strategy that is dynamically refined by client agreement throughout training. This approach allows AutoFLIP to identify robust and efficient sub-networks from the outset. Our extensive experiments show that AutoFLIP reduces computational overhead by an average of 52% and communication costs by over 65% while simultaneously achieving state-of-the-art accuracy in challenging non-IID settings.

cs.LG↗

HMACE: Heterogeneous Multi-Agent Collaborative Evolution for Combinatorial Optimization

Large Language Models have recently emerged as a promising paradigm for automated heuristic design for NP-hard combinatorial optimization problems. Despite this progress, existing LLM-based methods typically rely on monolithic workflows constrained by rigid templates, thereby restricting memory-guided exploration and triggering premature convergence to local optima. To design an autonomous and collaborative architecture, we introduce HMACE, a Heterogeneous Multi-Agent Collaborative Evolution framework that reconceptualizes heuristic search as an organizational design problem. HMACE decomposes each evolutionary generation into an autonomous, role-specialized loop with four coordinated agents: a Proposer for strategy exploration, a Generator for executable heuristic synthesis, an Evaluator for empirical assessment, and a Reflector for archive-backed memory update. By coupling behavior-aware retrieval, lightweight candidate filtering, and fitness-grounded archive updates, HMACE guides the search toward diverse and promising heuristic behaviors while avoiding redundant evaluations. Extensive evaluations on representative COPs, including TSP, Online BPP, MKP, and PFSP, show that HMACE achieves a favorable quality-efficiency trade-off compared to state-of-the-art single-agent and multi-agent baselines. In the matched LLM-driven reference comparison, HMACE achieves the lowest average gaps on TSP and Online BPP (0.464\% and 0.223\%, respectively), while requiring only 0.13M and 0.42M tokens for the two tasks, substantially fewer than the compared baselines.

cs.AI↗

Curriculum-guided multimodal representation learning enables generalizable prediction of nanomaterial-protein interactions

Nanomaterial-protein interactions (NPI) are pivotal to realizing the therapeutic and diagnostic potential of nanomaterials. Although AI promises to accelerate mechanistic understanding and enable rational nanomaterial design, robust generalization to unseen nanomaterials or proteins remains unresolved. Here, we present CuMMI (curriculum-guided multimodal interaction model), a generalizable, explainable, and transferable model designed to infer NPI across complex biological settings. CuMMI leverages a self-constructed million-scale NPI dataset and adopts a multi-stage curriculum centered on human plasma, with progressively broader biofluid exposure to enhance data coverage and generalizability. By integrating protein sequence, structure, and a text-encoded experimental context of 37 features, CuMMI captures complementary material-specific, biochemical, and environmental information. Sample-level quality weights are assigned to ensure full utilization of available data while mitigating low-confidence and sparsely recorded entries. Ablation studies highlight the most influential tabular features, clarifying their contribution to the prediction. Through rigorous external validation across independence-preserving temporal, nanomaterial-held-out, and protein-held-out evaluations, our framework consistently achieves good performance (mean of five classification metrics exceeding 0.75), highlighting its robustness and generalizability to unseen data. Furthermore, fine-tuning on independent gold-nanoparticle data and a held-out protein subset further delivers better performance than training from scratch with substantially fewer samples. Together, our approach enables generalizable and transferable NPI prediction and may accelerate in vitro research and applications of nanomaterials.

cs.LG↗

QuietWalk: Physics-Informed Reinforcement Learning for Ground Reaction Force-Aware Humanoid Locomotion Under Diverse Footwear

Humanoid robots operating in human-centered environments (e.g., homes, hospitals, and offices) must mitigate foot--ground impact transients, as impact-induced vibration and noise degrade user experience and repeated impacts accelerate hardware wear. However, existing low-noise locomotion training often relies on kinematic proxy objectives or fragile force sensors, and footwear-induced changes in contact dynamics introduce distribution shifts that hinder policy generalization.We present QuietWalk, a physics-informed reinforcement learning framework for ground-reaction-force-aware humanoid locomotion under diverse footwear conditions. QuietWalk employs an inverse-dynamics-constrained physics-informed neural network (PINN) to estimate per-foot vertical ground reaction forces (GRFs) from proprioceptive signals, and integrates the frozen predictor into the RL training loop to penalize predicted impact forces without requiring force sensors at deployment.On a held-out real-robot dataset, enforcing inverse-dynamics consistency reduces vertical GRF prediction errors by 82%-86% compared with a purely supervised predictor and improves the coefficient of determination from 0.39/0.67 to 0.99/0.99 for the left/right feet. On hardware at 1.2 m/s (barefoot; averaged over four floor materials), QuietWalk reduces mean A-weighted noise level by 7.17 dB and peak noise level by 4.98 dB under a consistent recording setup. Cross-footwear experiments (barefoot, skate shoes, athletic sneakers, and high heels) across multiple surfaces further demonstrate robust adaptation to footwear-induced contact variations.

cs.RO↗

Solver-Independent Automated Problem Formulation via LLMs for High-Cost Simulation-Driven Design

In the high-cost simulation-driven design domain, translating ambiguous design requirements into a mathematical optimization formulation is a bottleneck for optimizing product performance. This process is time-consuming and heavily reliant on expert knowledge. While large language models (LLMs) offer potential for automating this task, existing approaches either suffer from poor formalization that fails to accurately align with the design intent or rely on solver feedback for data filtering, which is unavailable due to the high simulation costs. To address this challenge, we propose APF, a framework for solver-independent, automated problem formulation via LLMs designed to automatically convert engineers' natural language requirements into executable optimization models. The core of this framework is an innovative pipeline for automatically generating high-quality data, which overcomes the difficulty of constructing suitable fine-tuning datasets in the absence of high-cost solver feedback with the help of data generation and test instance annotation. The generated high-quality dataset is used to perform supervised fine-tuning on LLMs, significantly enhancing their ability to generate accurate and executable optimization problem formulations. Experimental results on antenna design demonstrate that APF significantly outperforms the existing methods in both the accuracy of requirement formalization and the quality of resulting radiation efficiency curves in meeting the design goals.

cs.CL↗

OutSafe-Bench: A Benchmark for Multimodal Offensive Content Detection in Large Language Models

Since Multimodal Large Language Models (MLLMs) are increasingly being integrated into everyday tools and intelligent agents, growing concerns have arisen regarding their possible output of unsafe contents, ranging from toxic language and biased imagery to privacy violations and harmful misinformation. Current safety benchmarks remain highly limited in both modality coverage and performance evaluations, often neglecting the extensive landscape of content safety. In this work, we introduce OutSafe-Bench, the first most comprehensive content safety evaluation test suite designed for the multimodal era. OutSafe-Bench includes a large-scale dataset that spans four modalities, featuring over 18,000 bilingual (Chinese and English) text prompts, 4,500 images, 450 audio clips and 450 videos, all systematically annotated across nine critical content risk categories. In addition to the dataset, we introduce a Multidimensional Cross Risk Score (MCRS), a novel metric designed to model and assess overlapping and correlated content risks across different categories. To ensure fair and robust evaluation, we propose FairScore, an explainable automated multi-reviewer weighted aggregation framework. FairScore selects top-performing models as adaptive juries, thereby mitigating biases from single-model judgments and enhancing overall evaluation reliability. Our evaluation of nine state-of-the-art MLLMs reveals persistent and substantial safety vulnerabilities, underscoring the pressing need for robust safeguards in MLLMs.

cs.LG↗