arXiv:2503.12562v2 Announce Type: replace
Abstract: In Multiple Object Tracking (MOT), Re-identification (ReID) features are widely employed as a powerful cue for object association. However, they are often wielded as a one-size-fits-all hammer, applied uniformly across all videos through simple similarity metrics. We argue that this overlooks a fundamental truth: MOT is not a general retrieval problem, but a context-specific task of discriminating targets within a single video. To this end, we advocate for the adjustment of visual features based on the context specific to each video sequence for better adaptation. In this paper, we propose a history-aware feature transformation method that dynamically crafts a more discriminative subspace tailored to each video's unique sample distribution. Specifically, we treat the historical features of established trajectories as context and employ a tailored Fisher Linear Discriminant (FLD) to project the raw ReID features into a sequence-specific representation space. Extensive experiments demonstrate that our training-free method dramatically enhances the discriminative power of features from diverse ReID backbones, resulting in marked and consistent gains in tracking accuracy. Our findings provide compelling evidence that MOT inherently favors context-specific representation over the direct application of generic ReID features. We hope our work inspires the community to move beyond the naive application of ReID features and towards a deeper exploration of their purposeful customization for MOT. Our code will be released. The code is released at https://github.com/MCG-NJU/HATReID-MOT.
Science Journals
arXiv:2508.15017v2 Announce Type: replace
Abstract: Active Flux (AF) is a numerical method for hyperbolic conservation laws, whose degrees of freedom are averages/moments and (shared) point values at cell interfaces. It has been noted previously in a heuristic fashion that it thus combines ideas from Finite Volume/Discontinuous Galerkin (DG) methods with a continuous approximation common in continuous Finite Element (CG) methods. This work shows that semi-discrete Active Flux methods can be obtained from a variational formulation through a particular choice of (biorthogonal) test functions. These latter being discontinuous, the new formulation emphasizes the intermediate nature of AF between DG and CG. Explicit constructions are given for the case of arbitrarily high-order Active Flux with additional moments in 1-d, and for the classical third-order Active Flux on two-dimensional Cartesian meshes.
arXiv:2405.17838v3 Announce Type: replace
Abstract: Socio-linguistic indicators of affectively-relevant phenomena, such as emotion or sentiment, are often extracted from text to better understand features of human-computer interactions, including on social media. However, an indicator that is often overlooked is the presence or absence of information concerning harms or hazards. Detecting such indicators in text is important because substantial research demonstrates that negative events are more likely to be attended to, and more likely to elicit a response. In addition, statements about hazards are often found to be more believable than statements about benefits. Here, we develop a new model to detect information concerning hazards, trained on a new collection of annotated X posts. We show that not only does this model perform well, but that the hazard information it extracts is not strongly correlated with such widely used indicators as moral outrage, sentiment, and emotions. To demonstrate the utility of our tool, we apply it to two datasets of X posts that discuss important geopolitical events, namely the 2023 Israel-Hamas war and the 2022 French national election. In both cases, we find that hazard information, especially information concerning conflict, is common. We extract accounts associated with information campaigns from each data set to explore how information about hazards could be used to attempt to influence geopolitical events. We find that inorganic accounts representing the viewpoints of weaker sides in a conflict often discuss hazards to civilians, potentially as a way to elicit aid for the weaker side. Moreover, the rate at which these hazards are mentioned differs markedly from organic accounts, likely reflecting information operators' efforts to frame the given geopolitical event for strategic purposes. These results are first steps towards exploring hazards within an information warfare environment.
arXiv:2503.23536v3 Announce Type: replace
Abstract: Unlearnable data (ULD) has emerged as an innovative defense technique to prevent machine learning models from learning meaningful patterns from specific data, thus protecting data privacy and security. By introducing perturbations to the training data, ULD degrades model performance, making it difficult for unauthorized models to extract useful representations. Despite the growing significance of ULD, existing surveys predominantly focus on related fields, such as adversarial attacks and machine unlearning, with little attention given to ULD as an independent area of study. This survey fills that gap by offering a comprehensive review of ULD, examining unlearnable data generation methods, public benchmarks, evaluation metrics, theoretical foundations and practical applications. We compare and contrast different ULD approaches, analyzing their strengths, limitations, and trade-offs related to unlearnability, imperceptibility, efficiency and robustness. Moreover, we discuss key challenges, such as balancing perturbation imperceptibility with model degradation and the computational complexity of ULD generation. Finally, we highlight promising future research directions to advance the effectiveness and applicability of ULD, underscoring its potential to become a crucial tool in the evolving landscape of data protection in machine learning.
arXiv:2504.00617v5 Announce Type: replace
Abstract: Exceptional points (EPs) are central to non-Hermitian physics because of their unique properties and broad application prospects. While extensively studied in parity-time ($\mathcal{P}\mathcal{T}$)-symmetric systems and under Markovian dynamics, their exploration in broader pseudo-Hermitian settings, particularly in those involving non-Markovian environments, remains largely unexplored. In this study, we investigate a pseudo-Hermitian system consisting of three coupled optical cavities interacting with non-Markovian environments. Compared to the Markovian baseline, we demonstrate that the emergence of non-Markovian memory effects enlarges the dimensionality of the parameter space of the system, thereby giving rise to higher-order EPs. Furthermore, we show that these non-Markovianity-induced higher-order EPs admit a topological characterization as defects in the relevant pseudo-Hermitian parameter space, with quantized charges described by resultant winding numbers. Moreover, we observe that the pseudo-Hermitian system with an effective gain induced by coherent perfect absorption enables the higher-order EPs to be directly read out from the output spectrum. We also note that the non-Markovian mechanism for generating higher-order EPs is not restricted to pseudo-Hermitian systems, but can be extended to generic non-Hermitian quantum systems. Additionally, possible experimental implementations based on superconducting circuits are also discussed.
arXiv:2506.04141v2 Announce Type: replace
Abstract: The sequential structure of videos poses a challenge to the ability of multimodal large language models (MLLMs) to locate multi-frame evidence and conduct multimodal reasoning. However, existing video benchmarks mainly focus on understanding tasks, which only require models to match frames mentioned in the question (hereafter referred to as "question frame") and perceive a few adjacent frames. To address this gap, we propose MMR-V: A Benchmark for Multimodal Deep Reasoning in Videos. The benchmark is characterized by the following features. (1) Long-range, multi-frame reasoning: Models are required to infer and analyze evidence frames that may be far from the question frame. (2) Beyond perception: Questions cannot be answered through direct perception alone but require reasoning over hidden information. (3) Reliability: All tasks are manually annotated, referencing extensive real-world user understanding to align with common perceptions. (4) Confusability: Carefully designed distractor annotation strategies to reduce model shortcuts. MMR-V consists of 317 videos and 1,257 tasks. Our experiments reveal that current models still struggle with multi-modal reasoning; even the best-performing model, Gemini-2.5-pro, achieves only 64.3% accuracy. Additionally, current reasoning enhancement strategies (Chain-of-Thought and scaling test-time compute) bring limited gains. Error analysis indicates that the CoT demanded for multi-modal reasoning differs from it in textual reasoning, which partly explains the limited performance gains. We hope that MMR-V can inspire further research into enhancing multi-modal reasoning capabilities.
arXiv:2602.11961v3 Announce Type: replace
Abstract: Open large language models (LLMs) have demonstrated improving multilingual capabilities in recent years. In this paper, we present a study of open LLMs for multilingual machine translation (MT) across a range of languages, and investigate the effects of model scaling and data scaling when adapting open LLMs to multilingual MT through continual pretraining and instruction finetuning. Based on the Gemma3 model family, we develop MiLMMT-46, which achieves top-tier multilingual translation performance across 46 languages. Extensive experiments show that MiLMMT-46 consistently outperforms recent state-of-the-art (SOTA) models, including Seed-X, HY-MT-1.5, and TranslateGemma, and achieves competitive performance with strong proprietary systems such as Google Translate and Gemini 3 Pro. Models are released at https://huggingface.co/collections/xiaomi-research/milmmt-46. Codes are released at https://github.com/xiaomi-research/gemmax.
arXiv:2607.17907v1 Announce Type: cross
Abstract: The maximization of statistical complexity has long been associated with the emergence of probability distributions lying between perfect order and complete disorder. While previous studies have shown that complexity-maximizing distributions exhibit a two-level structure in finite discrete systems, an analogous unified treatment for both discrete and continuous probability spaces has remained unavailable. In this work, we develop a general variational framework for a generalized statistical complexity constructed from Shannon and Renyi entropies. We derive a common stationary equation governing both discrete probability masses and continuous probability densities and prove that every stationary solution necessarily possesses exactly two probability levels, establishing a universal core-halo structure. We further demonstrate that the optimization problem reduces to a single multiplicity parameter and prove that the global complexity maximum is attained by the smallest admissible core, corresponding to a single dominant state in the discrete case and an infinitesimal core in the continuous limit. These results provide a complete analytical characterization of the complexity-maximizing distributions and reveal a common mathematical structure underlying complexity optimization in both discrete and continuous settings. The framework establishes a unified foundation for generalized statistical complexity with potential applications in statistical mechanics, information theory, and the analysis of complex systems.
arXiv:2607.17909v1 Announce Type: cross
Abstract: The edge-of-chaos heuristic has long served as a guiding principle for designing reservoir computers, yet its relevance to machine performance remains elusive. Here, taking the spectral radius of the reservoir network as the control parameter, we show that the radius yielding the best forecasting performance does not coincide with the Lyapunov edge of the isolated, teacher-forced, or closed-loop generative reservoir. By analyzing the collective dynamics of the teacher-forced reservoir, we find that the target dynamics are represented mainly by stable Lyapunov modes whose finite-time stability is strongly modulated by the input. This finding motivates a stability-expressivity transfer index, which balances the stability of these modes against their expressivity in representing the target. Across chaotic and quasiperiodic targets, and for both asymmetric and symmetric reservoirs, this index accurately identifies the optimal spectral radius for autonomous forecasting.
arXiv:2601.03043v4 Announce Type: replace
Abstract: Large language models (LLMs) demonstrate strong capabilities across a wide range of complex tasks and are increasingly deployed at scale, placing significant demands on inference efficiency. Prior work typically decomposes inference into prefill and decode stages, with the decode stage dominating total latency. To reduce time and memory complexity in the decode stage, a line of work introduces sparse-attention algorithms. In this paper, we show, both empirically and theoretically, that sparse attention can paradoxically increase end-to-end complexity: information loss often induces significantly longer sequences, a phenomenon we term ``Less is Less'' (Lil). To mitigate the Lil problem, we propose an early-stopping algorithm that detects the threshold where information loss exceeds information gain during sparse decoding. Our early-stopping algorithm reduces token consumption by up to 90% with a marginal accuracy degradation of less than 2% across reasoning-intensive benchmarks.
arXiv:2607.17910v1 Announce Type: cross
Abstract: Generative artificial intelligence is rapidly transforming materials design by enabling de novo exploration of immense chemical spaces. Yet a large proportion of AI-generated compositions remain implausible, violating established chemical principles, which limits the reliability and interpretability of generative materials design. Here, we introduce a chemical validity operator that recasts heuristic chemical rules as a configurable algorithmic prior for evaluating and guiding generative materials discovery. Built on the open-source SMACT package, a data-informed oxidation-state model exposes tunable thresholds, allowing users to interpolate continuously between permissive and conservative chemical constraints, while supporting both exploratory and conservative materials-design workflows. Benchmarking six state-of-the-art generative models for inorganic crystals shows that most reproduce stoichiometry but under-represent realistic oxidation-state combinations, and that filtering removes compositions reliant on rarely observed oxidation states while preserving low-energy compounds near the convex hull. Beyond screening, the same operator can also serve as a reinforcement-learning reward, steering a latent diffusion model towards chemically grounded compositions. By encoding chemical heuristics and observations, this work establishes a foundation for oxidation-state-aware generative models.
arXiv:2607.17915v1 Announce Type: cross
Abstract: The present paper is a study of the large scale properties of the Kuramoto-Sivashinsky equation. By using a Schwinger-Dyson framework, we aim to provide a proof that the only solutions that can sustain a stable scaling in the infrared limit have a negative effective viscosity (in the Kuramoto-Sivashinsky sense) with a minimal set of hypotheses, thereby showing that this constitutes a general property of the wave equation that does not depend on a specific set of truncations of a renormalisation group flow, or limitations of a given numerical scheme for example.
arXiv:2512.19504v2 Announce Type: replace
Abstract: Cement production underpins global infrastructure but contributes approximately 7% of anthropogenic CO2 emissions, making accurate monitoring of production facilities essential for sustainable development. Existing remote sensing approaches rely predominantly on thermal signatures from kiln operations, which can be confounded by background heat sources and fail to capture persistent environmental alterations. This study introduces a physics-informed methodology that exploits multi-spectral features, particularly a geological Short Wave Infrared (SWIR) ratio, to detect soil property changes induced by sustained industrial heat emissions. This work proposes FusionNet, an intermediate multi-spectral data fusion framework that integrates Thermal Infrared (TIR) and SWIR inputs through a specialised backbone, embedding differential signal processing priors within a convolutional layer, mixed pooling, and wider receptive field. Systematic ablation studies confirm that each architectural component contributes to performance gains, with DGCNN achieving a 4.1-6.8% accuracy improvement over conventional CNNs. On the SWIR ratio dataset, FusionNet attains a maximum of 90.6%, outperforming state-of-the-art baselines across five spectral configurations and exceeding the strongest unimodal model by 1.1%. Transfer learning experiments reveal that ImageNet pretraining degrades TIR and SWIR performance, underscoring the importance of modality-aware training for cross-spectral applications. Overall, the results demonstrate that combining physics-aware feature selection with principled deep learning architectures enables robust, high-accuracy detection of cement production facilities, offering a reliable framework for industrial infrastructure monitoring
arXiv:2504.17356v3 Announce Type: replace
Abstract: Feature selection aims to preprocess the target dataset, find an optimal and most streamlined feature subset, and enhance the downstream machine learning task. Among filter, wrapper, and embedded-based approaches, the reinforcement learning (RL)-based subspace exploration strategy provides a novel objective optimization-directed perspective and promising performance. Nevertheless, even with improved performance, current reinforcement learning approaches face challenges similar to conventional methods when dealing with complex datasets. These challenges stem from the inefficient paradigm of using one agent per feature and the inherent complexities present in the datasets. This observation motivates us to investigate and address the above issue and propose a novel approach, namely HRLFS. Our methodology initially employs a Large Language Model (LLM)-based hybrid state extractor to capture each feature's mathematical and semantic characteristics. Based on this information, features are clustered, facilitating the construction of hierarchical agents for each cluster and sub-cluster. Extensive experiments demonstrate the efficiency, scalability, and robustness of our approach. Compared to contemporary or the one-feature-one-agent RL-based approaches, HRLFS improves the downstream ML performance with iterative feature subspace exploration while accelerating total run time by reducing the number of agents involved.
arXiv:2509.14335v2 Announce Type: replace
Abstract: Automated malware classifiers achieve strong detection performance, but auditing requires more than flagging a sample: analysts must explain malicious behaviors and justify them with code evidence. Traditional signature-based methods and learning-based XAI often fail to provide such support in a human-interpretable form. Large Language Models (LLMs) appear promising, yet their reliability for malware auditing remains unclear. Evaluation faces three challenges: (1) the lack of human-written behavioral ground truth; (2) real-world codebases that exceed current context limits; and (3) the lack of reliable mechanisms to verify whether generated claims are grounded in code evidence. These obstacles make benchmarking difficult and leave model capabilities and failure modes opaque.
We introduce MalEval, a diagnostic framework for measuring the capability boundaries of LLMs in malware auditing. MalEval pairs real-world application codebases with expert-written audit reports to provide fine-grained behavior-level ground truth. It compresses large codebases into behavior-relevant program contexts through a context-driven intermediate representation that preserves call relations. Expert reports and model outputs are mapped, via constrained reasoning, into structured evidence chains linking code-level facts to high-level behaviors in a shared space. MalEval decomposes auditing into four stage-wise tasks, enabling each intermediate judgment to be verified under limited context windows. We evaluate seven LLMs and find that they rely on surface cues rather than verifiable evidence, struggle to compose dispersed facts into coherent attack chains, and are highly sensitive to context formulation. These findings shift attention from isolated outputs to reliable LLM and agentic workflows for malware auditing. MalEval is publicly available at https://github.com/ZhengXR930/MalEval.git
arXiv:2505.02722v2 Announce Type: replace
Abstract: Although large language models (LLMs) have demonstrated impressive reasoning capabilities across general domains, their effectiveness in real-world clinical practice remains limited. This is likely due to their insufficient exposure to real-world clinical data during training, as such data is typically not included due to privacy concerns. To address this, we propose enhancing the clinical reasoning capabilities of LLMs by leveraging real-world clinical data. We constructed reasoning-intensive questions from a nationwide sepsis registry and fine-tuned Phi-4 on these questions using reinforcement learning, resulting in C-Reason. C-Reason exhibited strong clinical reasoning capabilities on the in-domain test set, as evidenced by both quantitative metrics and expert evaluations. Furthermore, its enhanced reasoning capabilities generalized to a sepsis dataset involving different tasks and patient cohorts, an open-ended consultations on antibiotics use task, and other diseases. Future research should focus on training LLMs with large-scale, multi-disease clinical datasets to develop more powerful, general-purpose clinical reasoning models.
arXiv:2607.16639v1 Announce Type: cross
Abstract: Biological systems across scales, along with many engineering problems, must control noisy systems with limited information. Here we study information-limited feedback control of stochastic systems to achieve target steady states, and derive a lower bound on the information rate from controlled system to controller. This framework allows us to obtain performance-information Pareto frontiers for wide-ranging control problems with limited information. For systems with state-independent passive dynamics, the bound is saturated by an explicit optimal control protocol which probabilistically time-reverses the passive dynamics. We showcase these results through applications to nonlinear particle localization, microbial navigation, and experimentally realized information engines.
arXiv:2505.06787v5 Announce Type: replace
Abstract: The algorithms developed for Maritime Autonomous Surface Ships (MASS) are often challenging to test on actual vessels due to high operational costs and safety considerations. Simulations offer a cost-effective alternative and eliminate risks, but they may not accurately represent real-world dynamics for the given tasks. Utilizing small-scale model ships and robotic vessels in conjunction with a laboratory basin provides an accessible testing environment for the early stages of validation processes. However, designing and developing a model vessel for a single test can be costly and cumbersome, and researchers often lack access to such infrastructure. To address these challenges and enable streamlined testing, we have developed an in-house testbed that facilitates the development, testing, verification, and validation of MASS algorithms in a digital-physical laboratory. This infrastructure includes a set of small-scale model vessels, a simulation environment for each vessel, a comprehensive testbed environment, and a digital twin in Unity. With this, we aim to establish a full design and verification pipeline that starts from low-fidelity and moves up to high-fidelity simulation models of each vessel, and thereby to the model-scale testing of the vessel in the laboratory basin. Further advancement allows moving towards semi-full-scale validation with R/V milliAmpere1 and full-scale validation with R/V Gunnerus. In this work, we present our progress on the development of this testbed environment and its components, demonstrating its effectiveness in enabling ship autonomy guidance, navigation, and control (GNC) algorithms.
arXiv:2505.10266v2 Announce Type: replace
Abstract: AI-generated misinformation (e.g., deepfakes) poses a growing threat to information integrity on social media. However, prior research has largely focused on its potential societal consequences rather than its real-world prevalence. In this study, we conduct a large-scale empirical analysis of AI-generated misinformation on the social media platform X. Specifically, we analyze a dataset comprising 82,076 misleading posts, both AI-generated and non-AI-generated, that have been identified and flagged through X's Community Notes platform. Our analysis yields four main findings: (i) AI-generated misinformation is more often centered on entertaining content and tends to exhibit a more positive sentiment than conventional forms of misinformation, (ii) it is perceived as less believable and less harmful than conventional misinformation, (iii) it more often originates from smaller user accounts, while authors posting such content are also associated with higher levels of partisanship and misinformation exposure, and (iv) AI-generated misinformation is significantly more likely to go viral. Altogether, our findings highlight the unique characteristics of AI-generated misinformation on social media and offer important implications for platforms and future research.
arXiv:2505.11491v3 Announce Type: replace
Abstract: This study investigates why physics-informed machine learning (PIML) can fail in macroscopic traffic flow modeling. We define failure as cases where a PIML model underperforms both purely data-driven and purely physics-based baselines by a given threshold. Unlike in other fields, physics residuals themselves do not hinder optimization in this setting. Instead, effective updates require both data and physics gradients to form acute angles with the true gradient, a condition difficult to satisfy with low-resolution loop data. In such cases, neural networks cannot accurately approximate density and speed, and the constructed physics residuals, already degraded by discrete sampling and temporal averaging, lose their ability to capture PDE dynamics, which directly leads to PIML failure. Theoretically, although LWR and ARZ solutions are weak solutions, for piecewise $C^k$ initial data they remain $C^k$ off the shock set under mild conditions, which has Lebesgue measure zero. Thus, almost all detector or collocation points lie in smooth regions where residuals are valid, and the MLP's inability to exactly represent discontinuities is immaterial. Finally, we establish MSE lower bounds of physics residuals: higher-order models such as ARZ have strictly larger consistency error bounds than LWR under mild conditions. This explains why LWR-based PIML can outperform ARZ-based PIML even with high-resolution data, with the gap shrinking as resolution increases, consistent with prior empirical findings.
arXiv:2505.11765v5 Announce Type: replace
Abstract: Agents powered by advanced large language models (LLMs) have demonstrated impressive capabilities across diverse complex applications. Recently, Multi-Agent Systems (MAS), wherein multiple agents collaborate and communicate with each other, have exhibited enhanced capabilities in complex tasks, such as high-quality code generation and arithmetic reasoning. However, the development of such systems often relies on handcrafted methods, and the literature on systematic design and optimization of LLM-based MAS remains limited. In this work, we introduce \textbf{OMAC}, a general framework designed for holistic optimization of LLM-based MAS. Specifically, we identify five key optimization dimensions for MAS, encompassing both agent functionality and collaboration structure. Building upon these dimensions, we first propose a general algorithm, utilizing two actors termed the Semantic Initializer and the Contrastive Comparator, to optimize any single dimension. Then, we present an algorithm for joint optimization across multiple dimensions. Extensive experiments demonstrate the superior performance of OMAC on diverse tasks against recent approaches.
arXiv:2505.17964v2 Announce Type: replace
Abstract: Cycle count statistics are fundamental tools in statistics and engineering, with applications in motif counting, channel coding, and statistical inference of network and matrix data. However, how to compute high-order cycle count statistics efficiently is still an open problem. In this paper, we aim to derive Computationally Efficient Equivalent Forms (CEEF) for cycle count statistics of any given order, where we express each cycle count statistic equivalently as a linear combination of finitely many terms. Using the CEEF, we provide a much more efficient way to compute the cycle count statistics.
The CEEF problem has no known general solution and requires delicate combinatorial arguments together with extensive calculations. While this task is hard to accomplish by humans alone, it provides an ideal setting in which Artificial Intelligence (AI) can be useful. We solve the problem by combining several theorems we derive with powerful coding skills of modern AI systems. Our results leverage graph-theoretic arguments and yield new formulas for general cases that were previously unknown. We find that, although AI cannot solve the problem independently, it becomes highly effective when guided by humans through theorems we derive as well as a clear derivation strategy, step-by-step instructions, and carefully-written prompts.
We consider several statistical applications, including spiked matrix testing, estimation of weak spike eigenvalues, and pairwise network comparison. For each problem, we demonstrate that optimal statistical performance is achieved by using high-order cycle count statistics, and our CEEF formulas make their computation feasible on large-scale data sets.
arXiv:2505.20161v2 Announce Type: replace
Abstract: Effective generalization in language models depends critically on the diversity of their training data. Yet existing diversity metrics often fall short of this goal, relying on surface-level heuristics that are decoupled from model behavior. This motivates us to ask: What kind of diversity in training data actually drives generalization in language models -- and how can we measure and amplify it? Through large-scale empirical analyses spanning over 300 training runs, carefully controlled for data scale and quality, we show that data diversity can be a strong predictor of generalization in LLM reasoning -- as measured by average model performance on unseen out-of-distribution benchmarks. We introduce G-Vendi, a metric that quantifies diversity via the entropy of model-induced gradients. Despite using a small off-the-shelf proxy model for gradients, G-Vendi consistently outperforms alternative measures, achieving strong correlation (Spearman's $\rho \approx 0.9$) with out-of-distribution (OOD) performance on both natural language inference (NLI) and math reasoning tasks. Building on this insight, we present Prismatic Synthesis, a framework for generating diverse synthetic data by targeting underrepresented regions in gradient space. Experimental results show that Prismatic Synthesis consistently improves model performance as we scale synthetic data -- not just on in-distribution test but across unseen, out-of-distribution benchmarks -- significantly outperforming state-of-the-art models that rely on 20 times larger data generator than ours. For example, PrismMath-7B, our model distilled from a 32B LLM, outperforms R1-Distill-Qwen-7B -- the same base model trained on proprietary data generated by 671B R1 -- on 6 out of 7 challenging benchmarks.
arXiv:2505.24852v3 Announce Type: replace
Abstract: On-device learning at the edge enables low-latency, private personalization with improved long-term robustness and reduced maintenance costs. Yet, achieving scalable, low-power end-to-end on-chip learning, especially from real-world sequential data with a limited number of examples, is an open challenge. Indeed, accelerators supporting error backpropagation optimize for learning performance at the expense of inference efficiency, while simplified learning algorithms often fail to reach acceptable accuracy targets. In this work, we present Chameleon, leveraging three key contributions to solve these challenges. (i) A unified learning and inference architecture supports few-shot learning (FSL), continual learning (CL) and inference at only 0.5% area overhead to the inference logic. (ii) Long temporal dependencies are efficiently captured with temporal convolutional networks (TCNs), enabling the first demonstration of end-to-end on-chip FSL and CL on sequential data and inference on 16-kHz raw audio. (iii) A dual-mode, multiplier-free compute array allows either matching the power consumption of state-of-the-art inference-only keyword spotting (KWS) accelerators or enabling $4.3\times$ higher peak GOPS. Fabricated in 40-nm CMOS, Chameleon sets new accuracy records on Omniglot for end-to-end on-chip FSL (96.8%, 5-way 1-shot, 98.8%, 5-way 5-shot) and CL (82.2% final accuracy for learning 250 classes with 10 shots), while maintaining an inference accuracy of 93.3% on the 12-class Google Speech Commands dataset at an extreme-edge power budget of 3.1 $\mu$W.
arXiv:2506.00625v2 Announce Type: replace
Abstract: The imbalanced distribution of long-tailed data presents a significant challenge for deep learning models, causing them to prioritize head classes while neglecting tail classes. Two key factors contributing to low recognition accuracy are the deformed representation space and a biased classifier, stemming from insufficient semantic information in tail classes. To address these issues, we propose permutation-invariant and head-to-tail feature fusion (PI-H2T), a highly adaptable method. PI-H2T enhances the representation space through permutation-invariant representation fusion (PIF), yielding more clustered features and automatic class margins. Additionally, it adjusts the biased classifier by transferring semantic information from head to tail classes via head-to-tail fusion (H2TF), improving tail class diversity. Theoretical analysis and experiments show that PI-H2T optimizes both the representation space and decision boundaries. Its plug-and-play design ensures seamless integration into existing methods, providing a straightforward path to further performance improvements. Extensive experiments on long-tailed benchmarks confirm the effectiveness of PI-H2T.