Papers

The effectiveness of active learning hinges on the choice of the acquisition criterion by which a learning algorithm selects potentially informative data points whose label is subsequently queried. This paper proposes a novel gradient-based acquisition criterion, derived from a generalization bound introduced by Luo et al. (2022). This criterion can be applied in lieu of uncertainty measures in uncertainty sampling, or incorporated into diversity-based methods that consider the spread of sampled points in addition to the uncertainty of their labels. We provide a theoretical justification of the proposed acquisition criterion, and demonstrate its effectiveness in an empirical evaluation.

Transformer-based models achieve state-of-the-art dependency parsing for high-resource languages, yet their advantage over simpler architectures in low-resource settings remains poorly understood. We evaluate four parsers—the Biaffine LSTM, Stack-Pointer Network, AfroXLMR-large, and RemBERT—across twelve typologically diverse languages, with a focus on low-resource African languages. We find that the Biaffine LSTM consistently outperforms transformer models in low-resource regimes, with transformers recovering their advantage as training data increases. The crossover falls within a resource range typical of treebanks for under-resourced languages. Morphological complexity (measured via MATTR) emerges as a significant secondary predictor of transformers' relative disadvantage after controlling for corpus size. These results indicate that the Biaffine LSTM may be better suited for syntactic tool development in low-resource regimes until sufficient annotated data is available to leverage the representational capacity of pre-trained transformers.

Fourier Neural Operators are deep learning models that learn mappings between function spaces and can be used to learn and solve partial differential equations (PDEs), in some cases significantly faster than traditional PDE solvers. Within the model are Fourier layers, which apply linear transformations directly to the Fourier modes, with parameters depending on the wave numbers. However, most physical systems are isotropic, with the results being independent of the coordinate system chosen, but the linear transformations do not necessarily respect these symmetries. We propose a modification to the linear transformations that ensures spatial symmetries are respected, called the Isotropic Fourier Neural Operator, which both improves model performance and reduces the number of parameters by up to a factor of 16 in 2D and 96 in 3D.

With each sensing modality exhibiting inherent strengths and limitations, multi-modal approaches for wearable Human Activity Recognition (HAR) are becoming increasingly relevant – particularly for recognizing Activities of Daily Living (ADLs), where individual modalities often produce ambiguous signals for similar or complex activities. This work introduces HARMES, a multi-modal wearable dataset combining three wrist-recorded modalities: motion sensing via an Inertial Measurement Unit (IMU), atmospheric environmental sensors (humidity, temperature, and pressure), and audio. Collected from 20 participants performing household activities in their own homes, HARMES totals over 80 hours of recorded data, with approximately three hours of labeled activity data per participant across 15 ADL classes. To the best of our knowledge, HARMES is the first dataset to combine this particular sensor trio, and it is nearly six times larger than the previously largest wrist-inertial-acoustic HAR dataset. In an extensive benchmark, we evaluate cross-subject generalization and conduct an ablation study revealing that modality contributions are activity-dependent and can provide complementary value, particularly for activities that are ambiguous from motion data alone. HARMES is freely available at Zenodo, alongside example code for loading the dataset and training models on GitHub.

Risk scores are an interpretable and actionable class of machine learning models with applications in medicine, insurance, and risk management. Unlike most computational methods, risk scores are designed to be computed by a human by attributing points to a data sample based on a limited set of criteria. The most common approaches for generating risk scores use linear regressions to estimate the effect of selected variables. We propose a simple and effective approach towards building compact and predictive risk scores. We provide an algorithm based on gradient boosting that is capable of modeling nonlinear effects, along with a C++ implementation with Python and R bindings. Through extensive empirical evaluation on twelve tabular datasets spanning regression, classification, and time-to-event tasks, we show that our method achieves competitive predictive performance while producing substantially more compact scores than regression-based alternatives, with 60% fewer rules for classification tasks and 16% fewer rules for time-to-event tasks on average, compared to AutoScore.

Optical Coherence Tomography (OCT) has become one of the most used imaging modality in ophthalmology. It provides high-resolution, non-invasive visualization of retinal microarchitecture. The automated analysis of OCT images through representation learning has emerged as a central research frontier. This has mainly been driven by the clinical need to process large acquisition volumes. The objective is to reduce the reliance on expert annotation, and improve diagnostic consistency across devices and populations. This survey provides a comprehensive and structured review of representation learning methods for retinal OCT image analysis. It covers the period from early deep learning approaches to the most recent developments in foundation models and vision-language systems. We organize the literature along a principled taxonomy of learning paradigms, encompassing supervised learning with CNN-based and transformer-based architectures, self-supervised and semi-supervised methods, generative approaches, as well as 3D volumetric modeling, multimodal representation learning, and large-scale pretrained foundation models. For each paradigm, we analyze the core methodological contributions, identify persistent limitations, and trace the connections between successive approaches. We further provide a structured overview of publicly available OCT datasets, discuss evaluation protocol considerations, and present a unified problem formulation that situates each learning paradigm within a common mathematical framework. Building on this analysis, we identify and discuss the most pressing open research directions emerging in the literature. This includes volumetric foundation model pretraining, uncertainty-aware representation learning, federated and privacy-preserving training, fairness and bias mitigation, concept-based interpretability,...

Large language models (LLMs) have shown promise as interactive agents that solve tasks through extended sequences of environment interactions. While prior work has primarily focused on system-level optimizations or algorithmic improvements, the role of task horizon length in shaping training dynamics remains poorly understood. In this work, we present a systematic empirical study that examines horizon length through controlled task constructions. Specifically, we construct controlled tasks in which agents face identical decision rules and reasoning structures, but differ only in the length of action sequences required for successful completion. Our results reveal that increasing horizon length alone constitutes a training bottleneck, inducing severe training instability driven by exploration difficulties and credit assignment challenges. We demonstrate that horizon reduction is a key principle to address this limitation, stabilizing training and achieving better performance in long-horizon tasks. Moreover, we find that horizon reduction is related to stronger generalization across horizon lengths: models trained under reduced horizons generalize more effectively to longer-horizon variants at inference time, a phenomenon we refer to as horizon generalization.

DeepSeek-V3.2 and V4 introduce Compressed Sparse Attention (CSA): a lightning indexer (a learned scoring projection over compressed keys) scores them, the top-k are selected per query, and a sparse attention kernel reads only those. Public CSA implementations materialize a [B, S, H_I, T] FP32 score tensor before the top-k reduction. With H_I=64 indexer heads and the V4-Flash compression ratio m=4, that intermediate is 256 GB at sequence length S=65,536, exceeding any single-GPU high-bandwidth-memory (HBM) budget. We present StreamIndex, a Triton implementation of the CSA pipeline whose central component is a chunked partition-merge top-k driver that never materializes the full intermediate. On synthetic-but-realistic V4-shaped inputs at the indexer-step (layer) level on a single NVIDIA H200, the materialize path runs out of memory (OOMs) at S=65,536 with V4-Flash dimensions; StreamIndex runs the same indexer to S=1,048,576 with 6.21 GB peak HBM, a 32x regime extension. Set-overlap recall against the materialize ground truth is bit-exact at small S where both fit; across three 5-point design-space sweeps (chunk size, key-tile size, top-k), mean recall rounds to 1.0000 with min recall at least 0.9980 in every cell. The chunked driver composes with TileLang's pipelined attention kernel: at S=262,144 with V4-Flash dimensions, the materialize indexer paired with TileLang attention OOMs while the chunked indexer paired with the same attention runs in 1.97 s at 18.56 GB peak. Our contribution targets the indexer step; we make no claim of a faster attention kernel or of real-checkpoint end-to-end behavior. Code: https://github.com/RightNow-AI/StreamIndex.

Chemotherapy dose optimization can be formulated as a dynamic treatment regime, requiring sequential decisions under uncertainty that must balance tumor suppression against toxicity. However, most reinforcement learning approaches assume full observability of the patient state, a condition rarely met in clinical practice. We investigate whether memory-augmented policies can improve chemotherapy control under partial observability. To this end, we employ a recurrent TD3-based approach with separate LSTM actor-critic networks and evaluate it on the AhnChemoEnv benchmark from DTR-Bench, considering both off-policy and on-policy recurrent architectures against feed-forward TD3 and Soft Actor-Critic. Pharmacokinetic and pharmacodynamic variability are held fixed to isolate hidden-state uncertainty and observation noise and to avoid confounding effects from inter-patient variability. Across ten random seeds, recurrence yields modest benefit under full observability but substantially stronger and more stable performance under partial observability, with more consistent tumor suppression and improved normal-cell preservation. These findings indicate that memory-based policies are particularly beneficial when clinically relevant state information is incomplete or noisy.

Deep reinforcement learning (DRL) navigation policies often overfit to the structure of their training environments, as environmental diversity is typically constrained by the manual effort required to design diverse scenarios. While procedural map generation offers scalable diversity, no prior work systematically compares how different generator types affect policy generalization. We integrate four generators (sparse, maze, graph, and Wave Function Collapse) with guaranteed navigability into MuRoSim, a 2D simulator focusing on training efficiency for LiDAR-based navigation. We cross-evaluate five navigation policies on 1000 seeded maps per generator across three training seeds. Results show a strongly asymmetric cross-generator transfer: a specialist trained on sparse layouts falls to 3.3% success on mazes, whereas a policy trained on the combined generator set achieves 91.5 +/- 1.1% mean success. We further demonstrate that A* path-planner subgoal inputs are the dominant factor for robustness, raising success from the 90.2 +/- 1.4% feedforward baseline to 98.9 +/- 0.4% and outperforming GRU recurrence, which only improves the reactive baseline. The DRL policies outperform a classical Carrot+A* controller, which matches their success only at low speeds (1.0 m/s) but collapses to 24.9% at 2.0 m/s. This highlights learned speed adaptation as the decisive advantage of the learned approach. Real-world experiments on a RoboMaster confirm sim-to-real transfer in a cluttered arena, while a maze-like layout exposes remaining failure modes that recurrence helps mitigate.

Accurate reconstruction of greenhouse climate variables from sparse sensor measurements is essential for intelligent environmental monitoring, automated climate control, and precision agriculture. In practical greenhouse operation, sensor failures, communication interruptions, calibration drift, and measurement noise frequently result in incomplete observations, making reliable estimation of indoor temperature and relative humidity a challenging inverse problem. This paper presents a coupled physics-informed neural network (PINN) for simultaneous reconstruction of greenhouse temperature and relative humidity and identification of unknown physical parameters governing a reduced greenhouse climate model. The framework integrates measurement data with coupled energy- and moisture-balance equations and initial-condition constraints, enabling climate state estimation and parameter identification within a unified learning framework. The methodology is evaluated using real greenhouse measurements under two validation protocols: random interpolation from sparse observations (Experiment A) and chronological temporal extrapolation over an unseen future interval (Experiment B). The proposed PINN is compared with a fully connected neural network, a long short-term memory (LSTM) network, and a gated recurrent unit (GRU) network. Under interpolation, the proposed PINN achieves the highest temperature reconstruction accuracy with an RMSE of $0.4495\,^{\circ}\mathrm{C}$ and an $R^2$ value of 0.9636, while simultaneously identifying physically interpretable model parameters. The two protocols provide complementary assessments of greenhouse climate reconstruction under interpolation and temporal extrapolation. The proposed framework provides a practical foundation for intelligent greenhouse monitoring, virtual sensing, digital twins, and automated greenhouse climate management.

Classic Network Intrusion Detection Systems (NIDS) often rely on manual feature engineering to extract meaningful patterns from network traffic data. However, this approach requires domain expertise and runs counter to the widely adopted principle of modern machine learning and neural networks: that models themselves should learn meaningful representations directly from data. We investigate whether tabular representation learning techniques can improve intrusion detection performance by automatically learning robust feature representations for NetFlow data. This paper presents a systematic evaluation of state-of-the-art representation learning methods on benchmark NetFlow datasets, comparing against traditional autoencoders and end-to-end transformer baselines. We evaluate learned representations using both supervised classifiers and unsupervised anomaly detectors, with comprehensive hyperparameter exploration for each combination. Our results reveal strong dataset-model dependency, with no single approach consistently dominating across all scenarios. For supervised classification, TabICL achieves the best performance on CIDDS, while autoencoders follow closely and tie with end-to-end transformer models for the best average rank across datasets. Supervised approaches substantially outperform unsupervised anomaly detection methods, where no single combination consistently dominates as optimal choices depend on the dataset. Cross-dataset transfer experiments demonstrate that learned representations can generalize across network environments with appropriate method and classifier selection. However, transfer performance varies substantially depending on the source-target dataset combination, indicating sensitivity to distributional differences between network environments.

Continual learning systems face a fundamental tension between plasticity – acquiring new knowledge – and stability – retaining prior knowledge. We introduce MPCS (Multi-Plasticity Continual System), a neuroplastic architecture that integrates eleven complementary mechanisms: task-driven neurogenesis, Fourier-encoded inputs, EWC regularization, meta-replay, mixed consolidation, hybrid gating, synapse pruning/regeneration, Hebbian updates, task similarity routing, adaptive growth control, and continuous neuron importance tracking. We evaluate MPCS on MEP-BENCH, a multi-track benchmark spanning 31 tasks across regression, classification, logic, and mixed domains, using a three-dimensional Pareto criterion over task performance (Perf), representation diversity (RD), and gradient conflict rate (GCR). Across 15 ablation configurations (3 seeds x 4 tracks x 2000 epochs), MPCS achieves a Normalized Efficiency Score of 94.2, placing it on the Pareto frontier among 9 of 14 gate-passing systems. Key findings: (i) Fourier encoding is the single most critical component (removal drops Perf by 30.7 pp and fails the MEP gate on 14% of tasks); (ii) global EWC degrades performance (NES = -4.2); topology-local EWC reduces this penalty (NES 90.5->91.8) but does not eliminate it; removing EWC entirely yields MPCS_EFFICIENT, the highest-Perf system – establishing a monotone relationship in the high task-similarity regime (s_bar ~= 0.95): global EWC < topology EWC < no EWC; (iii) the Pareto status assessment is predictive: removing the two Pareto-dominated components (EWC + Hebbian) jointly yields MPCS_EFFICIENT, which improves Perf by 0.6 pp at 4.7x lower compute cost (127 vs. 602 min), validating the Pareto frontier as an actionable model-compression guide.

Accurate prediction of Remaining Useful Life (RUL) in aero-engines is vital for predictive maintenance, improved operational reliability, and reduced lifecycle costs. While deep learning approaches have demonstrated strong potential in this area, most existing methods focus primarily on model architecture design and treat input features uniformly, often neglecting the influence of data preprocessing. In this work, we propose a novel preprocessing pipeline that enhances RUL prediction by improving data quality and temporal representation before model training. Our approach leverages complete temporal sequences and generates RUL estimates at each timestep, enabling the model to capture fine-grained degradation dynamics and deliver continuous prognostic insights throughout the engine's operational life. To validate the effectiveness of the proposed pipeline, we conduct experiments on the NASA C-MAPSS dataset. Comparative evaluations against a suite of state-of-the-art neural models including CNN, RNN, LSTM, DCNN, TCN, BiGRU-TSAM, AGCNN, and ATCN, demonstrate that our approach consistently achieves superior accuracy and robustness in aero-engine RUL prediction. These results highlight the critical role of preprocessing in maximizing the effectiveness of neural prognostic models.

Sleep foundation models have recently demonstrated strong performance on in-domain polysomnography tasks, including sleep staging, apnea detection, and disease risk prediction. In this work, we investigate whether sleep biosignals can serve as an effective pretraining distribution for learning representations that transfer beyond sleep to adjacent domains. Following sleep foundation models, we perform sleep-only multimodal contrastive pretraining (with a leave-one-out objective) and evaluate transfer to non-sleep EEG and ECG, two well-benchmarked biosignal modalities with heterogeneous datasets and clinically meaningful downstream tasks. Across eight downstream tasks spanning multiple EEG and ECG datasets, sleep pretraining consistently improves performance relative to training from scratch. Moreover, on several tasks, we achieve performance competitive with or surpassing prior specialized state-of-the-art and foundation models.

Offline Reinforcement Learning from Human Feedback (RLHF) pipelines such as Direct Preference Optimization (DPO) train on a pre-collected preference dataset, which makes them vulnerable to preference poisoning attack. We study label flip attacks against log-linear DPO. We first illustrate that flipping one preference label induces a parameter-independent shift in the DPO gradient. Using this key property, we can then convert the targeted poisoning problem into a structured binary sparse approximation problem. To solve this problem, we develop two attack methods: Binary-Aware Lattice Attack (BAL-A) and Binary Matching Pursuit Attack (BMP-A). BAL-A embeds the binary flip selection problem into a binary-aware lattice and applies Lenstra-Lenstra-Lovász reduction and Babai's nearest plane algorithm; we provide sufficient conditions that enforce binary coefficients and recover the minimum-flip objective. BMP-A adapts binary matching pursuit to our non-normalized gradient dictionary and yields coherence-based recovery guarantees and robustness (impossibility) certificates for $K$-flip budgets. Experiments on synthetic dictionaries and the Stanford Human Preferences dataset validate the theory and highlight how dictionary geometry governs attack success.

Online reinforcement learning with verifiable rewards (RLVR) turns checkable outcomes into a scalable training signal, but it keeps rollout generation, verifier scoring, and reference-policy evaluations on the optimization path. Static weighted supervised fine-tuning (SFT) on precomputed rollouts seems to remove this bottleneck, yet a weighted likelihood is not specified by rewards alone: its sampler and weights induce the policy being fit. This paper identifies the reference-sampled weighted-SFT objective whose induced policy equals the fixed-reference KL-regularized RLVR optimizer. The optimizer is the standard Boltzmann target policy, obtained by exponentially tilting the reference policy by verifier reward. Matching a weighted-SFT induced policy to this target forces density-ratio weights; in the reference-sampled subclass, this reduces uniquely, up to prompt scaling, to the prompt-normalized Boltzmann weight $\exp(r(x,y)/\beta)/Z(x)$. BOLT, a Boltzmann-Targeted SFT procedure, is the empirical estimator of this projection. The finite one-shot analysis separates the exact stored-support price $\beta\log(1/\pi^*(S_N\mid x))$ from partition estimation, effective-sample-size variance, generalization, optimization, and approximation errors. This decomposition explains why extra SFT epochs cannot repair missing reference-policy coverage and exposes the temperature–coverage–variance frontier. When coverage needs adaptive sampling, refreshed Boltzmann projections become KL policy mirror descent; finite inner solves enter as additive drift from the exact mirror step. Single-run Qwen experiments provide projection evidence for the target-matched weight, one-shot saturation, refreshed-sampler gains, and optimization-time savings, within the stated single-run scope.

We study the problem of black-box optimization of a function f of any dimension, given function evaluations perturbed by noise. The function is assumed to be locally smooth around one of its global optima, but this smoothness is unknown. Our contribution is an adaptive optimization algorithm, POO or parallel optimistic optimization, that is able to deal with this setting. POO performs almost as well as the best known algorithms requiring the knowledge of the smoothness. Furthermore, POO works for a larger class of functions than what was previously considered, especially for functions that are difficult to optimize, in a very precise sense. We provide a finite-time analysis of POO's performance, which shows that its error after n evaluations is at most a factor of sqrt(ln n) away from the error of the best known optimization algorithms using the knowledge of the smoothness.

Middle-mile logistics describes the problem of routing parcels through a network of hubs linked by trucks with finite capacity. We rephrase this as a multi-object goal-conditioned MDP. Our method combines graph neural networks with model-free RL, extracting small feature graphs from the environment state.

In this work, we formulate a new multi-task active learning setting in which the learner's goal is to solve multiple matrix completion problems simultaneously. At each round, the learner can choose from which matrix it receives a sample from an entry drawn uniformly at random. Our main practical motivation is market segmentation, where the matrices represent different regions with different preferences of the customers. The challenge in this setting is that each of the matrices can be of a different size and also of a different rank which is unknown. We provide and analyze a new algorithm, MAlocate that is able to adapt to the unknown ranks of the different matrices. We then give a lower-bound showing that our strategy is minimax-optimal and demonstrate its performance with synthetic experiments.

We present a machine learning framework for testing general relativity (GR) with gravitational wave signals from binary black hole mergers. Using the source parameters of 173 BBH events from the GWTC catalog as a realistic astrophysical population, we generate simulated GR waveforms and construct beyond GR (BGR) waveforms by applying controlled phase deformations. We introduce a response function formalism that provides a systematic framework for quantifying how any observable responds to modifications of GR. We train convolutional neural networks (CNNs) on two input representations: whitened waveforms and a response function type observable derived from the waveform mismatch, which isolates the effect of phase deviations from the bulk signal. Using response functions as the CNN input improves the classification sensitivity by a factor of approximately 33 compared to whitened waveforms, demonstrating that the choice of observable representation is as important as the classifier architecture. We study the fundamental limits of this classification through Bayes optimal error analysis, averaging methods that reveal coherent patterns hidden in noise, and a comparison between CNN accuracy and a single feature classifier as a proxy for human performance. At all deformation scales, the CNN outperforms the best single feature approach. We extend the framework to physically motivated theories using the parameterized post Einsteinian (ppE) formalism and apply it to massive gravity, where the classifier detects deviations for graviton masses of order $m_g \sim 10^{-23}\;\mathrm{eV}/c^2$ with aLIGO design sensitivity.

Encoder-decoder imbalance and the reliance on large input volumes make many 3D brain tumor segmentation models both compute-heavy and brittle. We present M\textsuperscript{4}Fuse, a lightweight network that prioritizes discriminative brain tumor cues over exhaustive appearance reconstruction. Our method balances encoder and decoder capacity and replaces depth expansion with a synergistic design: it propagates long-range context with linear complexity via a grouped state space mixer, denoises and aligns skip features using a cross-scale dual-stage gating bridge, and absorbs cross-site acquisition shifts with a sample-level mixture-of-experts. On the BraTS2019 and BraTS2021 benchmarks, M\textsuperscript{4}Fuse outperforms other lightweight excellent methods in both parameter count and performance. Even at a challenging input resolution of \(64\times128\times128\) (half that of existing excellent models), M\textsuperscript{4}Fuse reduces parameters by 62.63% and improves average performance by 0.09%. Ablations of key components validate the method's exceptional parameter-to-accuracy efficiency and robustness across diverse data centers.

Synthesizing realistic and diverse anomalous samples from limited data is vital for robust model generalization. However, existing methods struggle to reconcile fidelity and diversity, often hampered by distribution misalignment and overfitting, respectively.To mitigate this, we introduce Anomaly Preference Optimization,a novel paradigm that reformulates anomaly generation as a preference learning problem.Central to our approach is an implicit preference alignment mechanism that leverages real anomalies as positive references, deriving optimization signals directly from denoising trajectory deviations without requiring costly human annotation. Furthermore, we propose a Time-Aware Capacity Allocation module that dynamically distributes model capacity along the diffusion timeline,prioritizing structural diversity during highnoise phases while enhancing fine-grained fidelity in low-noise stages. During inference, a hierarchical sampling strategy modulates the coherencealignment trade-off, enabling precise control over generation. Extensive experiments demonstrate that significantly outperforms existing baselines,achieving state-of-the-art performance in both realism and diversity.

Open-set supervised anomaly detection (OSAD) aims to identify unseen anomalies using limited anomalous supervision. However, existing prototype-based methods typically model normal data via a unimodal Gaussian prior, failing to capture inherent multi-modality and resulting in blurred decision boundaries. To address this, we propose Mixture Prototype Flow Matching (MPFM), a framework that learns a continuous transformation from normal feature distributions to a structured Gaussian mixture prototype space. Departing from traditional flow-based approaches that rely on a single velocity vector, MPFM explicitly models the velocity field as a Gaussian mixture prior where each component corresponds to a distinct normal class. This design facilitates mode-aware and semantically coherent distribution transport. Furthermore, we introduce a Mutual Information Maximization Regularizer (MIMR) to prevent prototype collapse and maximize normal-anomaly separability. Extensive experiments demonstrate that MPFM achieves state-of-the-art performance across diverse benchmarks under both single- and multi-anomaly settings.

The Distributional Alignment Game framework provides a powerful variational perspective on Answer-Level Fine-Tuning (ALFT). However, standard algorithms for these games rely on estimating logarithmic rewards from small batches, introducing a systematic bias due to Jensen's inequality that can destabilize training. In this paper, we systematically resolve this structural estimation bias. First, we generalize the alignment game to arbitrary Bregman divergences, showing that for a family of geometries inducing polynomial rewards, we can construct provably exact and unbiased estimators using U-statistics. Second, for the canonical KL divergence game where an exact solution is impossible, we derive a globally robust minimax polynomial estimator that is provably optimal, achieving the fundamental statistical error limit of $\Theta(1/K^2)$, which we establish via the Ditzian-Totik theorem. Finally, we synthesize these two approaches to propose a novel Variance-Optimal Augmented Polynomial Optimization Program (AQP) Estimator, proving that by systematically reducing variance, our method achieves not only optimal bias but also provably accelerated game convergence, leading to more efficient and stable training with zero online computational overhead.

A recurring pattern in "reasoning without training" is that base LLMs already assign non-trivial probability mass to correct multi-step solutions; the bottleneck is locating these modes efficiently at inference time. Power sampling provides a principled way to bias decoding toward such modes by targeting p_theta(x)^alpha with alpha > 1, but practical approximations must account for future-dependent correction factors that determine which prefixes remain promising. We introduce Auxiliary Particle Power Sampling (APPS), a blockwise particle algorithm for approximating the sequence-level power target with a bounded population of partial solutions. APPS propagates hypotheses in parallel using proposal-corrected power reweighting and refines their survival through future-value-guided selection at resampling boundaries. This redistributes finite compute across competing prefixes rather than committing to a single unfolding path, while providing a direct scaling knob in the particle count and predictable peak memory. We instantiate the future-value signal with short-horizon rollouts and also study an amortized variant that replaces rollouts with a lightweight learned selection head. AMore broadly, APPS improves the accuracy–runtime trade-off of training-free decoding, further supporting the view that inference-time power approximation can recover gains often attributed to post-training.

Frequent handovers and finite satellite capacity make multi-user mobility management challenging in low Earth orbit (LEO) satellite networks. This paper develops a capacity-aware multi-user handover framework based on a double deep Q-network (DDQN) with a dueling architecture. Both training and evaluation employ sequential user equipment (UE) processing with shared satellite occupancy, while action masking excludes invisible or capacity-infeasible satellites. The state incorporates satellite visibility, achievable link quality, remaining capacity, previous association, short-term visibility, and time information. A fixed multi-criteria reward balances throughput, blocking, handover cost, connection persistence, and short-term visibility, while expert trajectories are used only for replay-buffer initialization. Simulations under different UE densities and satellite capacities show that the proposed framework maintains high average throughput per UE and low blocking rates with moderate handover frequency. Ablation results show that removing the remaining-capacity vector has only a limited effect on performance when sequential occupancy tracking and feasible-action masking are retained.

This paper proposes a spatial-temporal learning-based routing framework for dynamic low Earth orbit (LEO) satellite networks. The routing problem is formulated as a partially observable Markov decision process (POMDP), where each satellite independently selects its next hop using network-state observations and destination information. A graph attention network (GAT) extracts spatial features from the dynamic satellite topology and queue states, while a long short-term memory (LSTM) network captures their temporal evolution. The resulting representation is integrated into a deep Q-network (DQN) for destination-aware next-hop selection without a separate traffic-prediction stage. Simulations in a 45-satellite LEO constellation under normal and bursty traffic demonstrate packet delivery performance comparable to a graph neural network (GNN)-enhanced routing baseline, although the proposed method incurs higher delay and queue occupancy. Ablation studies further demonstrate the complementary benefits of spatial and temporal learning.

Efficient reasoning is not only a matter of shortening an answer trace; for tool-using agents, it also depends on whether the agent is reasoning over the right action space. As API ecosystems scale to tens of thousands of endpoints, the semantic gap between user requests and tool documentation makes this problem concrete: static retrieval from the initial query can fail before planning begins, and stronger planning alone cannot recover a missing tool. We study this problem as budgeted test-time retrieval and introduce FitText, a training-free framework that makes the tool interface revisable during execution by generating, refining, and evolving natural-language pseudo-tool descriptions as retrieval probes. FitText supports serial refinement, parallel exploration, and Memetic Retrieval, which adds evolutionary selection, local refinement, and tool memory to avoid redundant search. On StableToolBench (16,464 APIs), Memetic FitText reaches an 84.3% pooled pass rate, improving +26.7 points over static retrieval, +22.2 over Single-Pass, +23.2 over Re-Invoke, and +27.5 over Xu-style root refinement. It leads on every evaluated current model, with gains growing alongside model capability, and produces the largest improvements on ambiguous multi-tool tasks where dynamic re-retrieval restores correct candidates after early mistakes. At 40-way concurrency, parallel population execution keeps batched wall-clock at 1.01x Single-Pass despite the added search work.

Bayesian Optimization is an iterative method, tailored to optimizing expensive black box objective functions. Surrogate models like Gaussian Processes, which are the gold standard in Bayesian Optimization, can be inefficient for inputs with permutation symmetries, as the most common kernels employed are better suited for vector inputs rather than unordered sets of items. Motivated by this issue, we turn to permutation invariant Bayesian Optimization for well placement in Carbon Capture and Storage projects. The high fidelity black box simulator is instructed to operate wells under group control, giving rise to permutation symmetries within injector and producer groups that cannot be exploited with standard GP kernels. In this work, our main contribution is a novel Gaussian Process kernel (GP-Perm) that encodes permutation invariance by comparing sets through a stable divergence between their induced empirical representations, and can be combined with standard kernels for additional vector-valued inputs. As a learned invariant baseline, we also consider a Deep Kernel Learning model (DKL-DS) using the Deep Sets architecture to learn a permutation-invariant embedding. We evaluate the proposed methodology across 8 use cases, comprising seven synthetic benchmarks and one realistic CCS case study (Johansen formation)