Papers

Time series forecasting often suffers from over-smoothing, especially when future dynamics are multi-modal. Forecasts may follow the coarse trend of the observed future, but fail to preserve sharp changes, oscillations, turning points, and regime transitions that define plausible dynamic evolution. In this work, we revisit over-smoothing from the perspective of latent dynamical mode compression: under partial observation and single-realization supervision, multiple plausible future modes can be weakened, merged, or averaged during forecasting. Based on this view, we propose Dirichlet-Guided Group Forecasting (DGF), a mode-preserving forecasting framework that explicitly models multiple mode-conditioned predictive distributions and uncertainty over their selection probabilities. DGF uses a Dirichlet-guided hierarchical sampling mechanism and reward-based optimization to encourage forecasts that are accurate, dynamically consistent, and mode-distinct. Extensive experiments on real-world forecasting benchmarks show that DGF reduces over-smoothing while improving forecasting accuracy, diversity, and dynamical consistency.

Large language models (LLMs) are on the rise for accelerating scientific discovery, most recently in advanced tasks such as generating valid scientific hypotheses. Yet in many discovery settings, the goal is not to identify a single best hypothesis since validation can be noisy and expensive, and scientists benefit from a set of high-quality alternative hypotheses that hedge against downstream uncertainty for the best solutions. Nevertheless, commonly used evolutionary search recipes tend to prioritize optimization over exploration in hypothesis generation, and the resulting selection pressure during the search process leads to diversity collapse. Motivated by these limitations, we formulate hypothesis search as a sampling problem, where the objective is to efficiently produce diverse, high-quality hypotheses under a fixed validation budget. Building on this perspective, we propose \ours, an evolutionary framework inspired by the classical parallel tempering algorithm that searches hypotheses at multiple temperature levels and enables principled information exchange across temperatures to improve exploration without disrupting convergence. Across domains including molecular discovery, equation discovery, and algorithm discovery, our approach consistently improves both hypothesis quality and diversity under the same validation budget, and produces candidates that remain robust under more expensive downstream computational validations.

We present NOVA, an autonomous symbolic regression framework that identifies interpretable car-following and lane-change structures from raw trajectory data with minimal behavioral priors. Applied to 4,765,788 active driving observations from the NGSIM I-80 and US-101 datasets, NOVA's deterministic Rust-powered search engine evaluates over 10,000 candidate algebraic structures and identifies a compact two-term acceleration model under a forward-shifted rolling-mean prediction target. Evaluated under two complementary preprocessing pipelines, NOVA achieves $RMSE = 1.376 m/s^2$ ($R^2 = 15.57\%$) on the intent-forecasting benchmark, outperforming the best recalibrated symbolic-regression baseline (SR-LLM, PNAS~2025) by 0.135 m/s$^2$ in RMSE under an identical evaluation protocol. Across eight independent experiments, a single dominant nonlinear term emerges as a robust backbone of human car-following; a residual-guided extension further links the selected structure to an established psychophysical theory of collision avoidance. The discovered feature operators transfer zero-shot between freeway sites with under 3 pp $R^2$ loss. Extended to lane-change modelling within a multinomial logit framework, NOVA achieves 67.4% balanced accuracy under strict vehicle-ID holdout on 502 unseen drivers, surpassing existing lane-changing baselines by +29.8 percentage points on a three-class problem.

Generalization in emergent communication has largely focused on novel inputs or linguistic structures, yet the capacity for agents to communicate with strangers from strictly disjoint communities remains relatively unexplored. In this work, we formalize this capability as zero-shot mutual intelligibility (ZMI): successful communication between independently trained populations without prior exposure. Leveraging emergent sketching – in which agents communicate through sets of drawn strokes – as a visually grounded modality, we find that scaling the training population substantially improves ZMI across independent groups. Crucially, as we scale the population size, in-group communicative variation increases, preventing co-adaptation into homogeneity. Simultaneously, cross-group variation decreases, indicating a structural convergence toward a certain type of universality. Further analysis reveals that this universality is achieved through perceptual grounding: scaled populations increasingly anchor their emergent sketches on the objective visual resemblance of the target images. Together, these results position ZMI as a distinct axis of generalization in emergent communication and suggest a route toward socially interoperable artificial agents.

The asymptotic behaviour of Monte Carlo optimistic policy iteration (MC-O-PI) is a long-standing open question. When the model of the environment is unknown, as is common in practice, the only known condition that guarantees convergence to optimality is impractical. In its canonical form, this condition requires that the episodes used for policy evaluation be initialised uniformly over the entire state-action space. This paper strictly relaxes that requirement. Specifically, we prove that initial-visit MC-O-PI converges to optimality even when updates are uniform only over the actions within each state. This allows episodes to start in different states at arbitrary frequencies; a realistic implementation when the state space is large or unknown but the action space in each state is manageable. The proof departs from the classical analysis of Tsitsiklis whose central commutativity argument no longer applies when states are updated at different frequencies. Instead, we first show that the mean-field dynamics of MC-O-PI generate monotonically improving policies when updates are uniform over the actions in each state, and then prove that noise cannot consistently prevent this improvement by extending the lock-in argument of the combined stability-ODE method. This approach suggests a new way to study optimistic policy-iteration algorithms in general.

We propose a new optimization method, the Nyström-enhanced relaxed scalar auxiliary variable method (N-RSAV), which incorporates curvature information into the RSAV framework to accelerate convergence while preserving an unconditional modified energy dissipation law. Existing RSAV-based methods rely solely on first-order information and often suffer from slow convergence, particularly for ill-conditioned problems such as those arising in physics-informed neural networks (PINNs). To address this limitation, we design the linear operator in the RSAV scheme using approximate Hessian information obtained from a randomized low-rank Nyström approximation. To preserve the dissipation structure, we enforce positive semidefiniteness through eigenvalue truncation. Furthermore, we introduce an adaptive strategy that reuses the approximate Hessian based on the deviation between the original and modified energies, significantly reducing computational cost. We also provide a convergence analysis of the RSAV scheme with a general positive semidefinite operator under the Polyak-Lojasiewicz (PL) condition and establish corresponding convergence guarantees for N-RSAV under the PL condition and an additional convexity assumption. Numerical experiments on ill-conditioned problems with effectively low-rank structure, including convex quadratic problems and training of PINNs, demonstrate that the proposed methods achieve substantially faster convergence than conventional RSAV-based approaches.

Energy Dispersive X-ray (EDX) tomography in Scanning Transmission Electron Microscopy (STEM) enables 3D compositional and elemental mapping at the nanoscale, but its use is limited by restricted tilt ranges and low-dose conditions required to avoid beam damage. Limited-angle acquisition introduces missing-wedge artefacts such as elongation and anisotropic resolution, while noisy low-dose data further degrade reconstruction quality and quantitative reliability. Here, we introduce an unsupervised deep learning framework based on Deep Image Prior with total variation regularization (DIP-TV) for limited-angle STEM-EDX tomography. We extend it to a multi-channel formulation (DIPm-TV) that jointly reconstructs multiple elemental maps by exploiting spatial correlations. Using a synthetic 3-channel phantom, we show that the method compensates for severe missing-wedge artefacts corresponding to approximately $100^\circ$ of missing angular range under moderate noise, outperforming simultaneous iterative reconstruction technique and compressed sensing approaches. We apply the method to 3D chemical analysis of Ge-Sb-Te (GST) memory devices in virgin (as-fabricated) and SET (crystalline) operational states. Samples were prepared as cross-sectional focused ion beam lamellae and acquired under a limited-angle tilt range from $-40^\circ$ to $+40^\circ$ with $5^\circ$ steps and a dose of $2.0\times10^5$ $e^-/Ang^2$. The multi-channel approach enables voxel-by-voxel elemental reconstruction using only EDX signals without external structural priors such as high-angle annular dark-field imaging. The reconstructed volumes show near-isotropic spatial resolution and reveal compositional heterogeneities associated with device operation. This approach enables 3D chemical characterization in experimentally accessible sample geometries where conventional methods fail due to severe angular limitations.

Designing functional biological sequences requires navigating vast discrete spaces under strict evolutionary and biophysical constraints. Discrete Flow Matching (DFM) offers a generative framework over such spaces, but existing approaches rely on biologically uninformative couplings and offer limited flexibility for variable-length sequence generation and fine-grained control. We propose a structured coupling that encodes domain-specific preferences among sequence elements, biasing the source distribution toward plausible regions without modifying the flow objective or training procedure. Building on this, we introduce a latent edit-based rate parameterization that models variable-length generation via edit operations conditioned on a shared global latent, akin to a latent variable model, while remaining tractable. We further introduce a latent classifier-free guidance mechanism that steers generation coherently in continuous latent space, along with Dirichlet-prior temperature scaling for test-time control over edit operations. Our method achieves state-of-the-art performance across diverse biological sequence tasks, including density estimation, unconditional and conditional DNA sequence generation, and peptide sequence generation.

Recent developments in brain recording are driving a demand for machine learning tools capable of decoding the latent structure of large populations of neurons. In this paper, we provide a comprehensive survey that outlines the trajectory of Latent Variable Models (LVMs) from early state-space models to more recent deep generative models. We organize the literature into three closely related domains: (1) Single-Region Latent Dynamics, which includes models such as linear dynamical systems to more complex dynamics represented by Recurrent Neural Networks (RNNs) and Neural Ordinary Differential Equations (ODEs); (2) Multi-Region Communication, which employs probabilistic as well as subspace methods to study how information is transferred across different brain areas considering synaptic propagation delays and network connectivity; and (3) Behavior-Aligned Modeling, which seeks to disentangle neural activity related to task performance from other internal states via supervised or contrastive learning. This survey also includes large-scale neural foundation models, such as Transformers and diffusion models, that rely on large-scale pre-training for optimal performance across subjects. Finally, we conclude and discuss benchmarks, evaluation criteria, and open challenges, such as the ability to identify causal links or directionality of communication, to facilitate future research for bridging interpretable brain dynamics with reliable neural decoding.

Current reinforcement learning from human feedback (RLHF) methods primarily rely on scalar rewards from a trained reward model (RM). While effective, scalar rewards are often noisy and fail to capture fine-grained preference differences, whereas RM hidden states encode richer semantic and preference information. We introduce the representation-aware advantage estimation, which leverages RM hidden states and models them as auxiliary signals for better advantage estimation. Specifically, we propose the Graph-based Advantage Estimation (GraphAE), treat each sampled group as a graph, where nodes correspond to responses and edges capture their similarity in the RM hidden space. Then advantages are computed via graph propagation, enabling each sample to incorporate contextual information from its neighbors. GraphAE is lightweight and can be seamlessly integrated into existing group-based RL algorithms. We apply GraphAE to GRPO, GSPO and RLOO, and conduct extensive experiments on different models and benchmarks. Empirical results show consistent improvements across three benchmarks, with gains of up to + 6.3 on Arena-Hard-v0.1, + 8.27 on AlpacaEval 2.0, and + 0.22 on MT-Bench. These results demonstrate that leveraging RM representations leads to more sample efficient and robust RLHF.

Traffic prediction is fundamental to intelligent transportation systems and urban computing, yet many cities continue to suffer from traffic data scarcity due to limited sensor deployment and uneven urban development. Cross-city knowledge transfer has thus attracted increasing attention, enabling data-rich cities to assist data-scarce ones. However, centralized approaches raise privacy concerns, while existing federated methods struggle with pronounced spatiotemporal heterogeneity across cities. To address these challenges, we propose MoE-FedTP, a personalized federated cross-city spatiotemporal prediction framework based on lightweight Mixture-of-Experts (MoE) networks. MoE-FedTP first employs spatiotemporal neural networks to extract features from both source and target cities, then introduces a set of expert networks derived from different source cities through partial parameter sharing. A gating mechanism dynamically fuses the experts to capture diverse traffic dynamics, achieving fine-grained modeling of urban heterogeneity while preserving privacy. Experiments on four real-world traffic datasets show that MoE-FedTP consistently outperforms state-of-the-art cross-city and federated learning baselines, demonstrating its effectiveness in enhancing prediction accuracy for data-scarce cities.

Local deployment of large Mixture-of-Experts (MoE) models falls short of the service quality achieved in cloud-scale environments, even under low-concurrency workloads. We identify four key gaps in local MoE inference: reliance on capacity-reduced models (quantized, distilled, rerouted), inability to meet 30-second TTFT for long prefills (more than 12K), sub-baseline decode throughput (under 20 tokens/s), and poor concurrency under mixed prefill-decode and batched decode workloads. We present a CPU-GPU hybrid system that achieves cloud-level SLOs on dual-socket commodity CPUs and consumer GPUs by (1) stream-loading prefill (SLP), boosting prefill throughput to 1,200 tokens/s and enabling 32K prompts within 30 seconds; (2) distributed SLP (DSLP) with SmallEP expert parallelism, reaching 1,800 tokens/s and 45K prompts in 30 seconds on two RTX 5090s; (3) intra-node prefill-decode disaggregation with zero-copy shared weights and a dual-batch attention-MoE overlap scheme, sustaining concurrency with under 15 percent latency increase and 50 percent throughput gains; (4) an AVX-512-optimized FP8 GEMV kernel, enabling native CPU FP8 inference while delivering 4-5x lower CPU latency; and (5) fine-grained CPU parallelism that attains 28 tokens/s on INT4 DeepSeek-V3 and 21.5 tokens/s on intact FP8 V3. Evaluations show our system delivers cloud-level QoS for flagship MoE models on consumer CPU-GPU platforms, reshaping local deployment with intact, original-precision inference and enabling high-quality, cost-effective access without datacenter infrastructure.

Deploying large language models in user-facing systems requires efficient output safety filtering. Existing approaches typically rely on a separate moderation model applied after generation, which doubles inference cost and only detects violations after generation completes. We observe that the signal needed for moderation is already present in the model hidden states. Based on this, we train lightweight token-level probes that operate directly on internal activations, producing per-token safety scores that can be aggregated for both offline evaluation and online intervention. The probe reuses activations from the generator and requires no additional forward pass, enabling sub millisecond per-token safety checks inside the decoding loop. A probe applied to a single mid layer recovers most decisions of a strong guard model, acting as a low cost surrogate optimized for latency rather than accuracy. In streaming settings, it can halt or modify unsafe outputs before they are fully generated, replacing end of sequence moderation with continuous token level monitoring. Compared to post hoc and streaming guard models, our method achieves orders of magnitude lower compute overhead with minimal latency cost. We also provide a practical deployment recipe, including layer selection, aggregation strategy, probing frequency, and triggering thresholds. Finally, we show that the probe linear component corresponds to a direction in residual space, enabling both detection and activation steering at negligible cost.

Parameter-efficient fine-tuning of large language models (LLMs) can exhibit problematic memorization of individual training examples. Empirical privacy auditing (EPA) quantifies this risk by measuring realistic data leakage on membership inference (MI) or reconstruction attacks. A key challenge in EPA is designing “canary” examples that are mixed with the privacy-sensitive training data. We propose generating synthetic canaries via high-temperature sampling ($T \geq 0.8$) from LLMs, using prompts tailored to the privacy-sensitive training data. These canaries act as high-influence outliers, ensuring high identifiability and hence strong audits. Further, since the canaries are themselves non-private, they are inspectable and can be inserted with repetition without jeopardizing the privacy of the real data. An important use of models fine-tuned on privacy-sensitive data is the generation of synthetic data. This also comes with privacy risk. We introduce a powerful synthetic data audit based on fine-tuning an auxiliary model on the synthetic data. Auditing the auxiliary model for the original canaries then provides a strong estimate of the privacy leakage through the synthetic data. Finally, leveraging our strong auditing methodologies, we perform a systematic investigation into the interacting effects of model capacity and canary entropy on memorization.

Dynamic multi-resource allocation is a central problem in shared computing environments, where users' demands arrive sequentially and resources must be distributed fairly without knowledge of future demands. Existing methods emphasize fairness guarantees such as Sharing Incentive, Envy Freeness, and Dynamic Pareto Optimality, but often overlook system utility. Moreover, these fairness criteria are mutually incompatible, preventing strict enforcement of them at the same time. We propose a neural allocation mechanism that reconciles fairness with utility through multi-objective optimization during sequential rollout. We first formalize fairness in the dynamic setting via stepwise loss functions for Sharing Incentive, Envy Freeness, and Dynamic Pareto Optimality, enabling differentiable training. Leveraging non-wastefulness, we parameterized the solutions by constraining allocations to the subspace of demand while allowing elastic over-allocation when resources remain available. Empirical results demonstrate that our learned allocator achieves substantially higher utility at comparable levels of fairness, uncovering clear Pareto-frontier-like tradeoffs across metrics.

In time-series generation, existing approaches typically handcraft ortrain a separate model for each dataset, which hinders their scalability and fails to leverage shared temporal structures across domains. To address this fragmentation, we propose UPLOTS, a Unified, Prompt-guided Language model framework fOr constrained Time-Series Generation across diverse domains. Instead of building task-specific models, UPLOTS leverages a single pre-trained transformer backbone guided by learned constraint prompts, enabling on-demand generation with precise pattern control. One key innovation is our dynamic multi-dataset loss re-weighting and prompt-to-pattern mapping, which allows UPLOTS to internalize diverse temporal structures during training and conditionally generate them at inference. We evaluate UPLOTS on four real-world benchmarks and multiple constraint settings, including peak-period, calendar, load-level, and volatility patterns. Additional held-out constraint-combination and downstream forecasting experiments further demonstrate that UPLOTS generalizes beyond the original peak-pattern setting and improves data augmentation under scarce real-data regimes. Our code and baselines are available at github repo: https://github.com/cruiseresearchgroup/UPLOTS.

Text-attributed Graphs (TAGs) incorporate textual node attributes with graph structures to describe rich relational semantics. Recent efforts to integrate Graph Neural Networks (GNNs) and Large Language Models (LLMs) have shown promise for learning on TAGs, yet achieving well-aligned representations remains challenging. Prior studies largely rely on heuristics that perform coarse-grained matching. They lack sufficient constraints and ignore distributional alignment, leading to representation drift and limited generalization. Building on Energy-based Models (EBMs), we propose an Energy-based Representation Alignment (ERAlign) framework that projects GNN-encoded graph structure and LLM-derived text embeddings in a shared latent space to achieve distribution consistency. Concretely, layer-wise alignment is quantified by a distance metric and optimized via an EBM objective. By decreasing energy values, our framework yields well-aligned representations for downstream tasks. During training, we introduce Energy Discrepancy (ED) to avoid high sampling costs associated with intractable normalization. ED also carries theoretical guarantees of higher training efficiency and reduced energy landscape distortion. Empirical evaluations on eight TAG datasets demonstrate that ERAlign obtains state-of-the-art performance across varying levels of supervision and cross-task transfer scenarios.

DiffC provides a principled way to reuse pre-trained diffusion models for lossy compression, but its encoding and decoding procedures remain slow because they require many discretized forward and reverse steps. We study whether few-step generative models – Rectified Flow, Consistency Trajectory Models (CTM), and MeanFlow – can be cast as codecs within the same reverse channel coding (RCC) framework. The main challenge is that RCC requires posterior and shared distribution parameters, whereas these models do not explicitly parameterize intermediate conditional distributions. For Rectified Flow and MeanFlow, we use the equivalence between velocity parameterization and diffusion-style denoising parameterization to derive the quantities required by RCC. For CTM, which is distilled from EDM, we adopt the EDM noise parameterization together with local Gaussian approximations of the sender and shared distributions at intermediate states. This yields a proof-of-concept probabilistic formulation that enables compression with pre-trained few-step generative models without retraining. On low-resolution benchmarks, the resulting codecs reduce encoding and decoding time and improve realism in the low-bit-rate regime.

The financial market is a typical low signal-to-noise ratio (SNR) setting, which often destabilizes off-policy maximum-entropy methods like Soft Actor-Critic (SAC). Specifically, noisy state representations may produce unreliable Q-value estimates, and bootstrapping amplifies these errors, forming a failure mode we call the "Financial Entropy Trap". In this paper, we propose FPQC-SAC, an efficient and plug-and-play SAC variant that places a compact and bounded Parameterized Quantum Circuit (PQC) before the actor and critic networks to constrain feature propagation at the representation level, rather than filtering raw inputs or regularizing Q-values after bootstrapping. Notably, FPQC-SAC reduces the impact of extreme market fluctuations on Bellman target estimation, while trainable quantum entanglement preserves flexible cross-asset interactions. Empirical evaluations on real-world portfolio management tasks demonstrate that FPQC-SAC substantially enhances out-of-sample stability and cumulative returns by achieving a 66.89% relative gain in cumulative return over standard unconstrained SAC and outperforms the best continuous-control deep reinforcement learning baseline by approximately 27%. Open-source code is available at https://github.com/ZeyuLIU-UST/FPQC-SAC-main.

Semi-structured 2:4 sparsity is widely supported by modern accelerators, providing up to a 2x theoretical speedup. However, its strict 50% sparsity constraint often causes non-negligible accuracy degradation under post-training pruning. Meanwhile, existing relaxed sparsity formats either require specialized compiler support or introduce runtime overheads that limit end-to-end speedup. We propose Spense, a practical hybrid sparse-dense format that splits each weight matrix into a 2:4 sparse region and a dense region. This design relaxes the effective sparsity constraint while remaining compatible with existing high-performance sparse and dense GEMM libraries, avoiding both custom compiler support and input activation expansion. Building on this format, we introduce SpenseGPT, a one-shot post-training pruning method that produces sparse and dense regions. Notably, we show that selecting the right dense regions is important, and we devise two different strategies to choose them. Experiments on Qwen3-32B and Seed-OSS-36B demonstrate that our method achieves up to 1.2x end-to-end decoding speedup on B200 GPUs with FP8 precision, while preserving accuracy. To the best of our knowledge, this is the first one-shot pruning demonstration of real-world end-to-end LLM decoding speedup from semi-structured sparse tensor cores on recent GPUs such as B200s, while maintaining model quality.

Distributed machine learning (ML) is a key paradigm for today's large-scale artificial intelligence applications. As model inference arises as an important use case, faithful modeling of latency-sensitive collective communication has never been more important. Capturing the device architecture and modeling control and data paths at high fidelity is therefore a necessity today. Having a common, detailed representation for distributed ML infrastructure is also crucial. We revisit the promising open-source, community-driven simulator: ASTRA-sim. In this work, we identify limitations of the current ASTRA-sim simulator and augment it with new features. To this end, we enable fine-grained, high-fidelity simulation with a standardized infrastructure representation, opening new design space exploration opportunities. We propose the simulation at cache-line-sized load-store granularity, with a detailed graphics processing unit (GPU) execution model, to balance simulation scalability and fidelity. We also introduce InfraGraph, a standardized representation to capture distributed ML network infrastructure in detail. Using the updated ASTRA-sim 3.0 simulator, we showcase interesting design space explorations for designing optimized collective algorithms, network requirements, and GPU architectures.

Transformers achieve strong language modeling performance by providing direct token-to-token communication paths, but causal self-attention scales quadratically with context length. Recurrent and state-space models reduce this cost, yet compress history into sequentially updated fixed-size states. This paper studies a third primitive: a parallel content-addressed memory over causal successor records. The proposed Parallel Causal Associative Field (PCAF) writes local records from a context window into hash buckets, retrieves a bounded candidate set for the current query, forms a sparse cache distribution over successor tokens, and mixes that cache with a parametric local language model through a learned gate. The resulting model maintains sparse long-context access while avoiding a single fixed recurrent state bottleneck. We evaluate PCAF under full autoregressive pretraining on WikiText-103 and PG-19 using a distributed Google Cloud TPU v4-32 pod. At 303M parameters and context length T = 2048, PCAF-semantic reaches 36.31 perplexity on WikiText-103 and 52.45 perplexity on PG-19, compared with 47.49 and 53.84 for a matched dense Transformer. PCAF-semantic simultaneously processes 0.61-0.62M tokens/s across the TPU pod, versus 0.43M tokens/s for dense and local attention baselines. Supporting 41M-parameter multi-seed sweeps and single-GPU component ablations show that the associative cache, retrieval capacity, and learned gate materially affect the speed-quality trade-off.

Objective: Accurate classification of physiological signals in real-world deployments is challenged by sensor noise, motion artifacts, and distribution shifts between training and deployment data. Inference-time augmentation (ITA), which applies augmentations during inference rather than retraining, offers a simple, model-agnostic mechanism to improve robustness. However, ITA application to physiological signals has remained narrow in scope, relying on limited augmentation methods with fixed, unoptimized parameters. This work proposes a unified ITA framework to address that gap. Approach: The framework incorporates 13 augmentation methods spanning time-domain, amplitude-domain, frequency-domain, and artifact-injection transformations, with hyperparameters optimized via Bayesian optimization. We evaluate on atrial fibrillation (AF) detection from 30-second PPG signals using GPT-PPG and ResNet across five datasets comprising more than 400 patients and ${\sim}$9,800 hours of recording. Main results: Standard ITA consistently improved AUROC (up to 8.5% for GPT-PPG and 0.7% for ResNet) and AUPRC (up to 10.6% for GPT-PPG and 0.8% for ResNet). Selective ITA further reduced average FPR by up to 4.4% (GPT-PPG) and 1.3% (ResNet) on non-AF datasets. Significance: These findings establish ITA as a practical, model-agnostic approach for improving PPG-based AF classification reliability in deployment settings where retraining is not feasible, with broader applicability to physiological signal analysis.

We argue that forgetting is not confined to continual learning but is a general optimization phenomenon: during standard training, dominant mini-batch gradients suppress rare but useful update directions, causing short-term forgetting at every step. When such knowledge is never revisited, these losses compound into long-term forgetting-the classical failure mode of continual learning. We introduce FOGO, a scalable optimizer that continuously detects and resolves gradient interference across both regimes. FOGO spectrally orthogonalizes momentum updates to prevent dominant directions from monopolizing optimization, then stores representative past directions in a compact codebook memory built on random projection, where pairwise distances are provably preserved in low-dimensional space. At each step, conflicts between the current update and stored directions are resolved via lightweight orthogonal correction and lifted back through a proximal step, with minimal overhead and no data storage. Across class-imbalanced classification, continual visual learning under domain and class shifts, continual fine-tuning of LLaVA-7B, and GPT-2 pretraining, FOGO consistently improves convergence and knowledge retention, outperforming Adam and Muon.

Credit-card fraud detection is difficult because fraudulent transactions are rare, costly, and unevenly distributed. Strong gradient-boosted tree models already perform well on structured transaction data, so the value of another fusion method is not obvious. This paper examines whether Combinatorial Fusion Analysis (CFA), which searches over model subsets and rank-score fusion rules, can still add value on the IEEE-CIS Fraud Detection benchmark. Using a leakage-free 60/20/20 train/validation/test protocol, we evaluate 480 fusion configurations built from seven base classifiers. The best test-set result comes from diversity-weighted score fusion of Random Forest, XGBoost, and LightGBM (DEF WtScore), with AUC-ROC = 0.9405, AUPRC = 0.6699, and F1 = 0.6373. Bootstrap confidence intervals from 1,000 resamples show that the gains over the strongest single model exclude zero for all three metrics. CFA matches soft voting on AUC-ROC, improves AUPRC and F1, and outperforms stacking in this setting. A CTGAN augmentation experiment gives a negative result: synthetic fraud samples degrade both individual models and CFA. Overall, CFA is most useful here not as a way to combine every classifier, but as a validation-stage method for choosing a small, complementary subset and assigning diversity-aware weights.

On-policy distillation (OPD) has demonstrated strong empirical gains in enhancing complex reasoning in LLMs by aligning a student model with a teacher's predictive distribution over the student's own trajectories. An emerging variant, Privileged OPD, further strengthens this paradigm by employing a self-teacher model augmented with privileged information, such as oracle traces, to mitigate teacher-student capacity gaps while providing dense, answer-directed supervision. However, current methods treat privileged information as a monolithic imitation target, failing to disentangle locally reachable reasoning steps from future-conditioned oracle signals. Consequently, the student is encouraged to match a hindsight-biased distribution that often falls outside its local predictive support. This reachability mismatch incentivizes the student model to skip valid intermediate reasoning in favor of locally unsupported shortcuts. To resolve this, we introduce Anchored Residual On-Policy Distillation (AR-OPD), a dual-view framework that disentangles privileged supervision. Rather than enforcing strict full-view imitation, AR-OPD establishes a locally compatible anchor using a partially privileged teacher, isolating and injecting oracle foresight as a controlled residual to provide destination-directed guidance. Across diverse reasoning tasks, AR-OPD outperforms full privileged OPD by 2.3 points and SFT by 7.9 points. Crucially, this anchored residual mechanism reduces hindsight leakage by 21.7% and mitigates late-stage drift, yielding up to a 7.2-point advantage on challenging long-horizon trajectories exceeding 768 tokens.

This paper analyzes bidirectional random projections for ordinary least squares (OLS) regression under the fixed design setting. Let $(X,Y) \in \mathbb{R}^{n \times p} \times \mathbb{R}^n$ be a sample and $R \in \mathbb{R}^{n_1 \times n}, W \in \mathbb{R}^{p \times p_1}$ be two properly distributed random projections. We develop an expected excess loss bound for the OLS estimator built on $(WXR, WY)$. Compared to an established bound for OLS estimator built on $(XR, Y)$, the gap is approximately $O\left( p_1 + C \frac{1}{p_1} \right)$, where $C$ scales with $n_1/n$ and can be negative for small $n_1/n$. Its implications are confirmed by numerical results on real-world data.

As large language models (LLMs) continue to scale, it becomes increasingly challenging to grow model capacity under fixed computation budgets. We propose Path-Aligned Decompression Distillation (PADD), a framework for distilling knowledge from dense teachers without explicit routing into mixture-of-experts (MoE) students while learning high-quality routing policies. PADD organizes knowledge distillation into four stages in two phases: an initialization phase (Stage I) that builds diverse functionality in the student's experts through teacher neuron clustering and student-expert warmup, and a training phase (Stages II–IV) that integrates online adaptive distillation, path-refined policy optimization, and reward-augmented load balancing in a single training pipeline. Experiments on mathematical reasoning benchmarks demonstrate that PADD yields substantial gains over strong baselines at the same inference cost and that the MoE student can match or surpass its dense teacher. They also demonstrate effective teacher-to-student knowledge distillation and stable routing behavior.

Convergence-rate analysis for classifiers is often conducted under either Tsybakov margin or Massart margin. The former is a relatively weak condition that typically yields polynomial rates, while the latter is substantially stronger but can guarantee exponential rates. In this paper, we introduce a new condition, called Boltzmann margin, that bridges the gap between these two regimes. It is weaker than Massart margin, generally stronger than Tsybakov margin, and can imply many of their properties under suitable conditions. We apply Boltzmann margin to the analysis of kNN classifiers and establish the first near-exponential convergence rates for kNN classification. We also present extensions of the main results and provide numerical evidence supporting the main theoretical implications.

Learning Bayesian network (BN) structure from sparse discrete data is hard: when each instance records only a few variables, most variable pairs lack the joint observations needed for reliable scoring, and data-only methods recover little structure. Imperfect domain knowledge, expressible as a weighted directed knowledge graph (KG), is often available. KG-SoftMAP encodes such a KG as a finite-strength, confidence-weighted edge prior and maximizes a MAP objective that adds this logit-form prior to the BDeu score. With an informative but imperfect KG, KG-SoftMAP recovers partial directed structure even at observation rate rho=0.05, with directed F1 (DF1) of 0.19-0.32 across benchmarks. At higher observation rates within this sparse grid, DF1 reaches 0.44-0.66 at rho=0.20 and 0.46-0.64 at rho=0.40. Across the same three rates, KG-SoftMAP without the KG prior averages DF1 0.00, 0.19, and 0.21. Stress tests that corrupt, remove, or blur the KG signal, together with checks on LLM-extracted graphs beyond canonical benchmarks, show that recovery rises and falls with KG quality. On three real sparse educational datasets without ground-truth DAGs, we evaluate prediction, calibration, and KG-consistency. On Short Answer Feedback (SAF), KG-SoftMAP+VE reaches Fail-class F1 0.75 versus 0.78 for logistic regression while also providing an inspectable concept graph, calibrated Fail probabilities, and posterior queries from partially observed concept evidence. The remaining datasets sharpen the operating picture: weak heuristic KG signal leaves prediction unchanged, while an independent expert ontology moves the learned graph toward expert relatedness.