arXiv:2606.01366v1 Announce Type: new
Abstract: Learned plasma transport surrogates can match short horizon states while failing over long rollouts because charge accounting, density admissibility, and Poisson compatible field reconstruction are not enforced. We study this issue in a controlled nondimensional one dimensional drift diffusion Poisson benchmark with Dirichlet electrostatic potential boundaries and zero species wall fluxes. The benchmark is a conservation and rollout test, not a complete sheath wall model. We compare Conservative FluxNet, a structure preserving flux correction model with a conservative finite volume update and positivity aware limiting, against direct next state regressors, direct variants with Poisson recomputation, charge projection, and rollout training, and a classical conservative core without learned correction. The central result is that the classical finite volume core alone achieves near roundoff rollout error, so the paper is primarily about conservative discrete structure rather than learned closure. On the headline experiment, the conservative model achieves rollout MSE $7.35\times 10^{-9}$ versus $4.23\times 10^{1}$ for the unconstrained baseline, $2.53\times 10^{1}$ with Poisson recomputation, $6.72\times 10^{1}$ with charge projection, and $2.71\times 10^{1}$ with four step rollout training. Across $64$ prespecified configurations, it wins rollout mean squared error in $60/64$ cases despite winning one step mean squared error in only $19/64$. These results show that, for this controlled benchmark and comparison class, local conservative finite volume structure is more important than one step neural regression accuracy for stable autoregressive rollout.
Science Journals
arXiv:2606.01460v1 Announce Type: new
Abstract: Multi-pitch estimation (MPE) typically predicts which pitches are active in a mixture, but not which instrument or source produced them. This paper investigates a lightweight slot-attention framework for multi-instrument MPE (MI-MPE), where a mixture CQT is mapped to an unordered set of source-like pitch maps. The model uses permutation-invariant Hungarian matching to avoid fixed output semantics and treats the number of slots as an upper bound on the number of active sources. We further study two modular extensions: a self-supervised timbre encoder that provides training-time targets for slot-level timbre embeddings, and a polyphony branch that regularizes the pitch density of mixture- and slot-level predictions. Experiments show that Hungarian matching substantially improves instrument family decomposition on URMP. Stem-level prediction remains more challenging: timbre and polyphony supervision improve selected configurations, but do not consistently resolve source assignment. The results suggest that slot-based architectures are a promising direction for source-aware MPE, while highlighting the need to couple auxiliary musical cues to slot identity more carefully.
arXiv:2605.26092v4 Announce Type: replace
Abstract: The deployment of Large Language Models (LLMs) and Vision Transformers (ViTs) on edge devices is significantly constrained by memory limitations and the critical timing bottlenecks introduced by dense Multiply-Accumulate (MAC) arrays. In the ultra-low bit regime, logarithmic Power-of-Two (PoT) quantization provides a hardware-efficient alternative by replacing MAC operations with bit-shifts. However, the non-uniform exponential lattice is inherently limited by a \textbf{Low Angular Resolution Regime}, a structural flaw that becomes particularly pronounced at sub-4-bit thresholds, leading to a notable degradation of high-dimensional feature manifolds. To address this geometric limitation, we propose Geometric Orthogonal Residual Projection Quantization (GoQuant), an algorithm-hardware co-design framework. By formulating quantization as a dual-basis geometric projection, GoQuant adaptively synthesizes a higher-resolution residual lattice using strictly shift-and-add operations. Furthermore, its analytical solver offers a practical alternative to computationally intensive gradient-based optimization, reducing the full-model calibration time for LLaMA-2-7B to approximately 15 minutes. Extensive evaluations demonstrate GoQuant's applicability across modalities and its hardware efficiency. Under the 3-bit (W3/A16) constraint, it achieves a perplexity of 6.10 on LLaMA-2-7B, comparing favorably to conventional MAC-intensive baselines like AWQ without relying on asymmetric scaling, while maintaining competitive accuracy in 4-bit scenarios. At the silicon level, standard-cell RTL synthesis at a 28nm node indicates that GoQuant effectively mitigates the timing bottlenecks associated with dense multiplier trees. By flattening the combinational logic depth, our parallel shift-and-add datapath reduces the critical path delay to 0.35 ns.
arXiv:2606.01565v1 Announce Type: new
Abstract: Vision-Language Navigation in Continuous Environments (VLN-CE) poses a formidable challenge for autonomous agents, requiring seamless integration of natural language instructions and visual observations to navigate complex 3D indoor spaces. Existing approaches often falter in long-horizon tasks due to limited scene understanding, inefficient planning, and lack of robust decision-making frameworks. We introduce the \textbf{Hierarchical Semantic-Augmented Navigation (HSAN)} framework, a groundbreaking approach that redefines VLN-CE through three synergistic innovations. First, HSAN constructs a dynamic hierarchical semantic scene graph, leveraging vision-language models to capture multi-level environmental representations, from objects to regions to zones, enabling nuanced spatial reasoning. Second, it employs an optimal transport-based topological planner, grounded in Kantorovich's duality, to select long-term goals by balancing semantic relevance and spatial accessibility with theoretical guarantees of optimality. Third, a graph-aware reinforcement learning policy ensures precise low-level control, navigating subgoals while robustly avoiding obstacles. By integrating spectral graph theory, optimal transport, and advanced multi-modal learning, HSAN addresses the shortcomings of static maps and heuristic planners prevalent in prior work. Extensive experiments on multiple challenging VLN-CE datasets demonstrate that HSAN achieves state-of-the-art performance, with significant improvements in navigation success and generalization to unseen environments.
arXiv:2503.07325v2 Announce Type: replace
Abstract: Understanding and certifying the behavior of modern deep neural networks remains a fundamental challenge in reliable machine learning. We introduce a new class of data-dependent generalization bounds that apply directly to trained models, without any modification. In particular, we present an exactly computable bound that is non-vacuous across all evaluated networks, including ImageNet-scale models with 600M parameters. This this is the first work showing that meaningful generalization guarantees are achievable even for large, unaltered deep networks.
Our approach reveals that generalization is governed by the interaction between the trained model and the geometry of the data distribution. We decompose the generalization error into two interpretable components: a distributional complexity term, capturing how the data mass is distributed across the input space, and local model-behavior terms, capturing the network's behavior within individual regions. This joint dependence identifies where and why generalization gaps arise. Empirically, some components of our bound are highly predictive of the true test error, and the bound tightens when the partition aligns with the intrinsic data geometry, highlighting data-dependent local regularity as a key driver of generalization.
arXiv:2503.24183v3 Announce Type: replace
Abstract: The expansion of ride-sourcing services such as Uber and Lyft has reshaped urban transportation by offering flexible, on-demand mobility via mobile applications. Despite convenience, these platforms confront significant operational challenges, particularly vehicle rebalancing-strategic repositioning of a fleet of vehicles to address spatiotemporal mismatches in supply and demand. Inadequate rebalancing results in prolonged rider waiting times and inefficient vehicle utilization, but also leads to fairness issues, such as the inequitable distribution of service and disparities in driver income. To tackle these, we introduce continuous-state mean-field control (MFC) and mean-field reinforcement learning (MFRL) models with continuous repositioning actions. MFC and MFRL offer scalable solutions by modeling each vehicle's behavior through interaction with the vehicle distribution, rather than with individual vehicles. This mitigates the curse of dimensionality with respect to the number of agents, enabling coordination across large fleets with significantly reduced computational complexity and eliminating the need to retrain the model when fleet size changes. To ensure equitable service access across geographic regions, we integrate an accessibility constraint into models and derive rebalancing policies that strike a balance between high fulfillment of rider demand and fair coverage of vehicle supply. Extensive evaluation using data-driven simulation of Shenzhen demonstrates the efficiency and robustness of our approach. Remarkably, it scales to tens of thousands of vehicles, with training times comparable to linear programming rebalancing. Besides, our policies effectively explore the efficiency-equity Pareto front, outperforming conventional benchmarks across key metrics like fleet utilization, fulfilled requests, and pickup distance, while ensuring equitable service access.
arXiv:2601.20115v3 Announce Type: replace
Abstract: As the role of modern Graphics Processing Units (GPUs) becomes increasingly essential for several computing tasks, analyzing their past and current progress is paramount for determining future constraints on scientific research. This is particularly compelling in the Artificial Intelligence (AI) domain, where rapid technological advancements and fierce global competition have led the United States to recently implement export control regulations limiting international access to advanced AI chips. Consequently, this paper examines technical progress in NVIDIA datacenter GPUs from the mid-2000s through 2025. Our main results identify doubling times of 1.43 and 1.67 years for FP16 and FP32 dense operations, while FP64 doubling times range from 2.05 to 3.79 years. Off-chip memory size and bandwidth have grown at slower rates than computing performance, doubling every 3.29 to 3.41 years, whereas the release prices and power consumption roughly doubled every 5.03 and 15 years, respectively. Moreover, our cross-vendor comparison of the top-performing GPUs per year shows that NVIDIA's performance advantage is narrowing, but not enough to compel a major market shift. Finally, we quantify the potential implications of current U.S. export control regulations and the consequent performance gaps, which the recently proposed policy changes could shrink from 23.6X to 3.54X.
arXiv:2602.05395v2 Announce Type: replace-cross
Abstract: A simple strategy for improving LLM accuracy, especially in math and reasoning problems, is to sample multiple responses and submit the answer most consistently reached. In this paper we leverage Bayesian prior information to save on sampling costs, stopping once sufficient consistency is reached. Although the exact posterior is computationally intractable, we further introduce an efficient "L-aggregated" stopping policy that tracks only the L-1 most frequent answer counts. Theoretically, we prove that L=3 is all you need: this coarse approximation is sufficient to achieve asymptotic optimality, and strictly dominates prior-free baselines, while having a fast posterior computation. Empirically, this identifies the most consistent (i.e., mode) LLM answer using fewer samples, and can achieve similar answer accuracy while cutting the number of LLM calls (i.e., saving on LLM inference costs) by up to 50%.
arXiv:2603.28439v2 Announce Type: replace
Abstract: Robots are increasingly being deployed in agriculture to support sustainable practices and improve productivity. They offer strong potential to enable precise, efficient, and environmentally friendly operations. However, most existing path-following controllers focus solely on the robot's center of motion and neglect the spatial footprint and dynamics of attached implements. In practice, implements such as mechanical weeders or spring-tine cultivators are often large, rigidly mounted, and directly interacting with crops and soil; ignoring their position can degrade tracking performance and increase the risk of crop damage. To address this limitation, we propose a closed-form predictive control strategy extending the approach introduced in [1]. The method is developed specifically for Ackermann-type agricultural vehicles and explicitly models the implement as a rigid offset point, while accounting for lateral slip and lever-arm effects. The approach is benchmarked against state-of-the-art baseline controllers, including a reactive geometric method, a reactive backstepping method, and a model-based predictive scheme. Real-world agricultural experiments with two different implements show that the proposed method reduces the median tracking error by 24% to 56%, and decreases peak errors during curvature transitions by up to 70%. These improvements translate into enhanced operational safety, particularly in scenarios where the implement operates in close proximity to crop rows.
arXiv:2606.01584v1 Announce Type: new
Abstract: Conversational tutoring agents have been shown to improve learning engagement and student outcomes, and large language models (LLMs) are increasingly used in these systems to provide scalable, personalized feedback. However, LLMs may perpetuate or amplify stereotypical social biases, posing particular risks in educational settings.
In this study, we evaluate LLMs in conversational tutoring scenarios to identify high-confidence social biases, instances where models are unable to identify biased judgments in tutoring conversations while maintaining strong confidence in their assessments, potentially affecting their reasoning and the feedback they provide to learners. We present a new dataset generation method that enables bias evaluation under naturalistic instructional conditions by regenerating student-AI tutor interactions and introducing turns with controlled bias derived from a benchmark dataset.
Using this data, we assess multiple LLMs' ability to detect stereotypical biases and analyze the confidence and reasoning underlying their responses through computational and human evaluations. We find that bias detection is substantially more challenging in conversational tutoring contexts than in benchmark-based evaluations, and that state-of-the-art LLMs are overconfident in their incorrect assessments of stereotypical bias statements. Moreover, model confidence strongly influences reasoning and feedback, highlighting the risks of overconfident, biased behavior in LLM-based tutoring agents. We conclude by discussing implications, mitigation considerations, and directions for future research.
arXiv:2504.04718v2 Announce Type: replace
Abstract: Recent studies have demonstrated that test-time compute scaling effectively improves the performance of small language models (sLMs). However, prior research has mainly examined test-time compute scaling with an additional larger model as a verifier, leaving verification by sLMs underexplored. In this work, we investigate whether sLMs can reliably verify the output candidates under test-time scaling. We find that even with knowledge distillation from larger verifiers, sLMs struggle with verification tasks requiring memorization, such as numerical calculations and fact-checking. To address this limitation, we propose Tool-integrated verification (T1), a two-stage framework that first filters candidates with external tools and then uses an sLM for final verification, offloading memorization-heavy steps to tools such as a code interpreter. Within T1, we prove that offloading to external tools reduces the memorization burden on sLMs and improves test-time scaling performance. Experiments on the MATH benchmark demonstrate that, with T1, a Llama-3.2 1B model under test-time scaling outperforms the significantly larger Llama-3.1 8B model. Moreover, T1 improves the verification accuracy of both process reward models (PRMs) and critic models. Our findings highlight the potential of tool integration to substantially improve the verification abilities of sLMs.
arXiv:2606.01594v1 Announce Type: new
Abstract: This paper considers a class of aggregative congestion games with uncertain coupling constraints, and devises a distributed algorithm to seek the robust generalized Wardrop equilibrium (RGWE) under worst-case uncertainty. Utilizing robust optimization theory, we reformulate the original aggregative congestion game with uncertainty into a tractable and deterministic augmented problem. Building upon this reformulation, we design a fully distributed algorithm to seek the RGWE by integrating a projected primal-dual scheme and a dynamic tracking technique. The convergence of the proposed algorithm is rigorously guaranteed via singular perturbation theory and LaSalle's invariance principle. Furthermore, we explicitly characterize the relationship between the obtained RGWE and the robust generalized Nash equilibrium, as the latter captures full strategic interactions. Finally, numerical simulations on the charging control of plug-in electric vehicles corroborate our theoretical findings.
arXiv:2606.01800v1 Announce Type: new
Abstract: Large language models (LLMs) have excelled in processing multiple languages through pre- and post-training on multilingual data, even though English dominates the training data. Prior work focusing on token representations has revealed how those LLMs process non-English text. Although these analyses have provided insightful findings, they fail to capture a structural view, which is an inherent property of language. In this study, we explore the multilinguality of LLMs through representational structural analysis. Our findings reveal that low-resource languages are structurally more different from English than high- and mid-resource languages, and that language-specific post-training alters their structures while preserving inter-language relationships.
arXiv:2606.01656v1 Announce Type: new
Abstract: Two fundamental questions have puzzled scientists for more than 150 years. How did life become homochiral and why was this specific handedness selected. Recently, it has been shown that homochirality could have emerged through the enantioselective interactions of molecules with magnetic substrates due to the asymmetric crystallization of an RNA precursor on a magnetite substrate, abundant on early Earth. This phenomenon is based on the chirality-induced spin selectivity,CISS, effect. Despite its robustness, this model could not provide an answer to the second question: why one specific handedness, D for RNA, was selected. Here we demonstrate that spin-involving processes can have different outcomes in the two enantiomers of chiral molecules. In chiral molecules with unpaired electrons or while electrons are passing through them, the total angular momentum vector, J, is aligned along the easy axis, which is defined by the magnetic anisotropy induced by the spin-orbit coupling and asymmetry of the molecular field. The magnitude J is the same for both enantiomers, but the vectors may be aligned differently relative to the molecular frame in the two enantiomers. This difference can be quantified by, for example, by the angle between J and electric dipole moment of the molecule. We show by direct measurements, theory, and ab initio calculations that dynamic spin processes in chiral molecules could result in different efficiencies of spin-related phenomena, including the interaction of chiral molecules with magnetic surfaces. The findings may provide an explanation for the specific homochirality in nature.
arXiv:2606.01660v1 Announce Type: new
Abstract: Pre-propagation graph neural networks (PPGNNs) push all graph-dependent computation into a preprocessing step and train only on the resulting dense hop features, which makes them highly scalable. A puzzle in this regime is that more complex hop aggregators do not reliably outperform simpler ones: on many benchmarks, a plain MLP-based aggregator matches or beats hop-attention variants. We revisit this behavior from a graph-filter perspective. Over a precomputed diffusion basis, existing PPGNNs differ mainly in how filter coefficients are shared across nodes and feature channels, rather than simply in raw aggregator capacity. MLP-based architectures learn channel-dependent filters that are largely shared across nodes, while hop-attention-based architectures learn node-dependent mixtures that are largely shared across channels. This reveals a missing regime in standard PPGNN designs: joint node- and channel-adaptive filtering under the pre-propagation computational contract. We propose FilterMoE, a mixture-of-experts PPGNN in which a small bank of learnable Chebyshev filter experts is routed jointly over nodes and channels by a 3D gating tensor. Across eleven homophilic and heterophilic benchmarks, FilterMoE outperforms strong PPGNN baselines on nine datasets and ranks first on all three large-scale benchmarks, improving the average test score by 1.53 points. These results establish joint node-channel filter routing as a robust alternative to dataset-specific hop-aggregator selection.
arXiv:2606.01676v1 Announce Type: new
Abstract: In response to the growing reliance on Agile and DevOps methodologies for enhancing software delivery speed and quality, this study investigates the persistent challenges and viable solutions associated with their integration. Although Agile promotes iterative development and customer responsiveness, and DevOps emphasizes automation and operational efficiency, their convergence in practice often presents significant organizational, structural, and technical hurdles. This research employs a qualitative methodology grounded in semi-structured interviews with six seasoned industry professionals across Brazil and Germany, each with extensive experience in both Agile and DevOps domains. The study identifies four core categories of integration challenges: Cultural & Organizational Barriers, Structural Constraints, Process \& Method Complexity, and Technical Limitations. Additionally, it offers four major solution domains: Team Structure & Autonomy, Culture & Collaboration, Process & Change Management, and Automation & Infrastructure. The findings underscore the importance of cultural alignment, proactive monitoring, automation, and other practices in mitigating integration friction. The results contribute to a deeper understanding of the Agile-DevOps interface and offer practical insights for software organizations seeking to navigate this complex transition effectively.
arXiv:2507.08158v2 Announce Type: replace
Abstract: Differential Privacy (DP) bounds the privacy leakage of a mechanism against worst-case membership inference, but the precise tradeoff between complex adversarial models and DP protections remains poorly understood. In this paper, we present a unified framework that generalizes the patchwork of existing bounds across membership inference, attribute inference, and data reconstruction attacks. Crucially, our framework is the first to evaluate attacks that target multiple individuals simultaneously and measure success beyond exact matches under a single cohesive bound. Our bounds capture this broad family of previously unexplored attack settings by relying solely on the privacy parameters and the adversary's baseline success rate (i.e. its prior without access to the mechanism's output). To illustrate this, we compare our high-probability guarantees to empirical attacks in two novel settings: extracting multiple non-uniform secrets (passwords and PII) from DP-finetuned language models, and reconstructing tabular data from noisy marginals. Ultimately, this framework provides a rigorous theoretical foundation to investigate the risk landscape of DP algorithms in new adversarial settings.
arXiv:2507.13833v4 Announce Type: replace
Abstract: Effectively scaling Reinforcement Learning (RL) is crucial for enhancing the reasoning and alignment of Large Language Models. The massive data and complex execution flows inherent in these tasks require a distributed architecture capable of efficient scaling. However, to simplify programming and dependency management, mainstream frameworks often rely on a centralized architecture where a single node dispatches both control and data. This inherent coupling creates significant communication bottlenecks, severely limiting system scalability and efficiency. We present DISTFLOW, a novel, fully distributed RL framework that adopts a multi-controller paradigm. By decoupling data transmission from control dispatch, DISTFLOW establishes a parallelism-aware, decentralized Data Coordinator that leverages local caching, load balancing, and asynchronous double buffer to minimize communication overhead and mitigate straggler effects. For control logic, it introduces a task scheduler built upon Directed Acyclic Graph (DAG) that facilitates fine-grained, independent execution. Experimental results demonstrate that DISTFLOW achieves near-linear scalability up to 512 GPUs and delivers up to a 2.63x throughput improvement over state-of-the-art (SOTA) frameworks. The source code is available at: https://github.com/sii-research/siiRL.
arXiv:2606.01679v1 Announce Type: new
Abstract: Multimodal LLMs are increasingly used to assist scientific peer review, where a core requirement is verifying whether claims in a paper are supported by its evidence. Prior work has shown that models perform substantially better at this task when the evidence is a table than when it is a chart of the same underlying data. This raises the question of whether models fail to extract information from charts, or do they extract it but fail to use it when forming their prediction? We study this question through layer-wise linear probing and attention analysis on three open-weight VLMs over table and chart evidence, representing the same underlying data. We find consistent evidence for the latter. Chart information is encoded in the models' intermediate representations but does not reach the prediction position, a gap that is absent for tables and holds across all conditions tested. Attention analysis further reveals that this disconnect takes two architecturally distinct forms across model families. These findings reframe the table-chart gap as a failure of how encoded visual information is routed at prediction time, rather than a failure of encoding itself.
arXiv:2606.01680v1 Announce Type: new
Abstract: Network failures are among the most frequent hardware faults in large-scale GPU clusters and a leading cause of training-job interruptions. Modern collective communication libraries such as NCCL mitigate network failures by rerouting traffic through surviving NICs on the same server, trading reduced inter-node bandwidth for uninterrupted training. However, the degraded server remains on the critical path of the standard ring algorithm, slowing the entire collective. We present the first information-theoretic lower bound on AllReduce completion time under asymmetric network bandwidth and show that when the straggler retains at least half of its original bandwidth, the unavoidable overhead relative to the fault-free optimum is only O(1/p) for p GPUs. We then design OptCC, a four-stage pipelined AllReduce algorithm that approaches this lower bound. Experiments on SimAI confirm that OptCC closes the gap left by existing fault-tolerant schemes: under practical network failures with up to 50% bandwidth loss, OptCC completes AllReduce within 2-6% of NCCL's fault-free ring performance, whereas the state-of-the-art incurs up to 57% overhead.
arXiv:2507.15336v3 Announce Type: replace
Abstract: Designing high-performance neural networks for new tasks requires balancing optimization quality with search efficiency. Current methods fail to achieve this balance: neural architectural search is computationally expensive, while model retrieval often yields suboptimal static checkpoints. To resolve this dilemma, we model the performance gains induced by fine-grained architectural modifications as edit-effect evidence and build evidence graphs from prior tasks. By constructing a retrieval-augmented model refinement framework, our proposed M-DESIGN dynamically weaves historical evidence to discover near-optimal modification paths. M-DESIGN features an adaptive retrieval mechanism that quickly calibrates the evolving transferability of edit-effect evidence from different sources. To handle out-of-distribution shifts, we introduce predictive task planners that extrapolate gains from multi-hop evidence, thereby reducing reliance on an exhaustive repository. Based on our model knowledge base of 67,760 graph neural networks across 22 datasets, extensive experiments demonstrate that M-DESIGN consistently outperforms baselines, achieving the search-space best performance in 26 out of 33 cases under a strict budget.
arXiv:2507.18863v2 Announce Type: replace
Abstract: Visual Automatic Speech Recognition (V-ASR) is a challenging task that involves interpreting spoken language solely from visual information, such as lip movements and facial expressions. This task is notably challenging due to the absence of auditory cues and the visual ambiguity of phonemes that exhibit similar visemes-distinct sounds that appear identical in lip motions. Existing methods often aim to predict words or characters directly from visual cues, but they commonly suffer from high error rates due to viseme ambiguity and require large amounts of pre-training data. We propose a novel phoneme-based two-stage framework that fuses visual and landmark motion features, followed by an LLM model for word reconstruction to address these challenges. Stage 1 consists of V-ASR, which outputs the predicted phonemes, thereby reducing training complexity. Meanwhile, the facial landmark features address speaker-specific facial characteristics. Stage 2 comprises an encoder-decoder LLM model, NLLB, that reconstructs the output phonemes back to words. Besides using a large visual dataset for deep learning fine-tuning, our PV-ASR method demonstrates superior performance by achieving 17.4% WER on the LRS2 and 21.0% WER on the LRS3 dataset.
arXiv:2508.03556v4 Announce Type: replace
Abstract: Process Reward Model (PRM) is widely used in the post-training of Large Language Model (LLM) because it can perform fine-grained evaluation of the reasoning steps of generated content. However, most PRMs lack long-term reasoning and deep thinking capabilities. On the other hand, although a few works have tried to introduce Chain-of-Thought (CoT) capability into PRMs, the annotation cost of CoT-PRM data is too expensive to play a stable role in various tasks. To address the above challenges, we propose VRPRM, a process reward model via visual reasoning, and design an efficient two-stage training strategy. Experimental results show that using only 3.6K CoT-PRM Supervised Fine-Tuning(SFT) data and 50K non-CoT PRM Reinforcement Learning (RL) training data, VRPRM can surpass the non-thinking PRM with a total data volume of 400K and achieved a relative performance improvement of up to 118\% over the base model in the BoN experiment. This result confirms that the proposed combined training strategy can achieve higher quality reasoning capabilities at a lower data annotation cost, thus providing a new paradigm for PRM training with more efficient data utilization.
arXiv:2508.06588v3 Announce Type: replace
Abstract: Vector Quantization (VQ) has recently emerged as a promising approach for learning compressed and discrete representations for graph-structured data. However, a fundamental challenge, i.e., codebook collapse, remains underexplored in the graph domain, significantly limiting the expressiveness and generalization of graph tokens.In this paper, we present an empirical study and observe that codebook collapse consistently occurs when training VQ jointly with Graph Neural Networks under graph reconstruction tasks, even with mitigation strategies proposed in vision or language domains. Moreover, we provide a diagnosis of collapse from data and optimization perspectives, showing that collapse is associated with graph data properties such as feature redundancy and connectivity density, and is further reinforced by the training dynamics of deterministic hard assignment. To address these issues, we propose RGVQ, a novel framework that integrates graph topology and feature similarity as explicit regularization signals to enhance codebook utilization and promote token diversity. RGVQ introduces soft assignments via Gumbel-Softmax reparameterization, ensuring that all codewords receive gradient updates. In addition, RGVQ incorporates a structure-aware contrastive regularization to penalize assigning the same token to dissimilar node pairs. Extensive experiments demonstrate that RGVQ substantially improves codebook utilization and consistently boosts the performance of state-of-the-art graph VQ backbones across multiple downstream tasks, enabling more expressive and transferable graph token representations.
arXiv:2508.10312v2 Announce Type: replace
Abstract: Recommender systems in concert with Large Language Models (LLMs) present promising avenues for generating semantically-informed recommendations. However, LLM-based recommenders exhibit a tendency to overemphasize semantic correlations within users' interaction history. When taking pretrained collaborative ID embeddings as input, LLM-based recommenders progressively weaken the inherent collaborative signals as the embeddings propagate through LLM backbones layer by layer, as opposed to traditional Transformer-based sequential models in which collaborative signals are typically preserved or even enhanced for state-of-the-art performance. To address this limitation, we introduce FreLLM4Rec, an approach designed to balance semantic and collaborative information from a spectral perspective. Item embeddings that incorporate both semantic and collaborative information are first purified using a Global Graph Low-Pass Filter (G-LPF) to preliminarily remove irrelevant high-frequency noise. Temporal Frequency Modulation (TFM) then actively preserves collaborative signal layer by layer. Note that the collaborative preservation capability of TFM is theoretically guaranteed by establishing a connection between the optimal but hard-to-implement local graph fourier filters and the suboptimal yet computationally efficient frequency-domain filters. Extensive experiments on four benchmark datasets demonstrate that FreLLM4Rec successfully mitigates collaborative signal attenuation and achieves competitive performance, with improvements of up to 8.00\% in NDCG@10 over the best baseline. Our findings provide insights into how LLMs process collaborative information and offer a principled approach for improving LLM-based recommendation systems.