Group Relative Policy Optimization(GRPO) has become a cornerstone of modern reinforcement learning alignment, prized for its efficacy in foregoing an explicit value-critic by leveraging reward normalization across sampled trajectory cohorts. However, the method's reliance on a monolithic statistical baseline, such as the group mean, collapses the relational topology of the trajectory space into a single scalar, thereby erasing the fine-grained preference information essential for navigating complex, rank-sensitive reward landscapes. To address this issue, we introduce a novel framework, Lambda Policy Optimization (LambdaPO), that addresses this information-theoretic bottleneck by re-conceptualizing advantage estimation from a scalar value to a decomposed, pairwise preference structure. Specifically, the advantage for any given trajectory is formulated as the integrated sum of reward differentials against all peers in its cohort, where each pairwise comparison is dynamically attenuated by the policy's own probabilistic confidence in the established preference. To further mitigate the sparsity of binary outcome supervision, we augment the objective with a semantic density reward, derived from the precision-recall alignment between generated reasoning traces and ground-truth solutions. As a result, our method can mine more fine-grained optimization signals from a group of rollouts, guiding the LLM to a better optima. Experimental results across challenging math reasoning and question-answering tasks demonstrates that LambdaPO improves performance compared to the baseline methods.

Short-form video poses new challenges to the quality assessment of user-generated content (UGC) due to its complex generation pipeline, rapid content variation, and mixed distortions. To address this challenge, we propose an end-to-end video quality assessment (VQA) framework that employs a dense visual encoder based on CLIP, and incorporates compression priors derived from the frequency domain to generate artifact- and structure-aware weight maps for feature aggregation. By explicitly decomposing artifact, structure, and original visual feature branches and adaptively fusing them over time through a learned gating module, the proposed method achieves accurate and efficient quality prediction. Experimental results show that our method achieves strong performance on short-form video datasets in terms of average rank and linear correlation (SRCC: 0.736, PLCC: 0.787), while maintaining efficient inference runtime. The code and additional results are available at: https://github.com/xinyiW915/FGSVQA.

Nowadays, quantum communications provide a vast field of research in rapid expansion, with a huge potential impact on the future developments of quantum technologies. In particular, continuous variable systems, employing coherent-state encoding and quadrature measurements, represent a suitable platform, due to their compatibility with both the modulation and detection systems currently employed in standard fiber-optical communications. In this work, we address some relevant aspects of the field, and provide innovative results being also experimentally oriented. In particular, we focus on two relevant paradigms: quantum decision theory and continuous variable quantum key distribution (CVQKD). In the former case, we address the problem of coherent-state discrimination and design new hybrid receivers for binary phase-shift keying discrimination, obtaining a quantum advantage over conventional detection schemes, being also robust against typical experimental imperfections. In the latter scenario, we proceed in two different directions. On the one hand, we design new CVQKD protocols employing discrete modulation of coherent states, being a feasible solution compatible with the state of the art in optical communications technologies. On the other hand, we address the more fundamental problem of performing channel losses mitigation to enhance existing protocols, and investigate the role of optical amplifiers for the task. Finally, we make a first step towards a fully non-Gaussian CVQKD scheme by proposing, for the first time, the adoption of an optimized state-discrimination receiver, commonly adopted for quantum decision theory, within the context of CVQKD, obtaining a genuine quantum enhancement over conventional protocols.

While optimal transport (OT) enforces a rigid constraint by requiring two measures to be matched exactly, partial optimal transport relaxes this requirement by allowing mass to remain unmatched through a global budget, scalar rebate, or uniform rejection rule. However, many applications call for more structured, pointwise rejection mechanisms, where the decision to leave mass unmatched depends on side-specific reliability, support geometry, or external information about which components should participate in the comparison. We introduce \emph{intent-controlled partial optimal transport} (IC-POT), a targeted generalization of partial transport that replaces the global rejection paradigm with pointwise rejection costs over both measures. We show that the resulting optimization problem admits a dual interpretation in terms of local acceptance thresholds and can be solved by recasting it as a balanced Kantorovich OT problem on an augmented support. Beyond theoretical analysis, we demonstrate the practical relevance of IC-POT in settings where rejection is driven by side information. In positive-unlabeled learning and open-partial domain adaptation, incorporating pointwise rejection rules that encode statistical structure improves fixed baseline pipelines. Finally, we motivate the use of IC-POT with a geophysical practical case: multi-modal satellite ocean measurements, for which physical and sensors priors naturally inform the rejection mechanism and define the retrieved comparable signal information.

Thermal cycle environments involving repeated temperature changes are common conditions observed in modern engineering processes. Under such conditions, materials undergo repeated thermal expansion and contraction, forming complex thermal stress fields. Thermal-crystal plasticity simulations that account for stress fields and thermal conduction at the polycrystalline microstructure scale are an effective method for numerically reproducing thermal cycle environments. However, the influence of thermal cycle frequency on the temporal behavior of the stress field and plastic response has not yet been fully understood, partly because a systematic analysis method capable of simultaneously capturing spatial heterogeneity and temporal evolution remains limited. In this study, we predicted the thermal stress field generated under different thermal cycle frequencies using thermal-crystal plasticity simulations and investigated the effect of frequency on the spatiotemporal structure of the stress response. The present framework illustrates that the resulting thermal-mechanical response can be represented as a superposition of multiple effective temporal components, reflecting the increased complexity of the system behavior. By employing dynamic mode decomposition (DMD) as a diagnostic and post-processing technique, we demonstrate that the spatiotemporal structure of the stress field under thermal cycle conditions can be systematically extracted and compactly represented. This approach enables a quantitative characterization of frequency-dependent changes in the thermal stress response beyond conventional averaging or snapshot-based analyses. The results highlight the utility of DMD as a framework for organizing complex simulation data and for interpreting the temporal structure of plastic response under cyclic thermal loading.

Discrete Gaussian Sampling on lattices is a fundamental problem in lattice-based cryptography. It appears both in basic cryptographic primitives such as digital signatures and as an important cryptanalysis building block for solving hard lattice problems. In this paper, we show a quantum algorithm based on the quantum rejection sampling technique whose complexity is asymptotically quadratically faster than its classical counterpart in [Wang & Ling, IEEE Trans. Inf. Theory 2019]. Our sampler outputs a quantum state which can either be measured to get the desired distribution or be used directly as such in other quantum algorithms. By doing so, we derive two versions of quantum dual attacks that improve upon the previous ones in [Pouly & Shen, EUROCRYPT 2024]. The two versions are incomparable, each having distinct advantages (speed vs memory requirement). The second version is particularly interesting as it requires only polynomial classical and quantum memory, excluding the classical memory used in the preprocessing step of the Discrete Gaussian sampler. Our quantum Discrete Gaussian sampler can also be used to speed up the algorithm for solving the Short Integer Solution problem, in any norm, of [Bollauf, Pouly & Shen, ePrint 2026/225].

The rapid growth of autonomous driving datasets has enabled the scaling of powerful motion forecasting models. While large-scale pretraining provides strong performance, the standard imitation objective may not fully capture the complex nuances of human driving preferences. Meanwhile, recent advances in vision-language models (VLMs) have demonstrated impressive reasoning and commonsense understanding. Building on these capabilities, this paper presents VL-DPO, a vision-language-guided framework that aligns ego-vehicle motion forecasting models with human preferences. Our approach leverages a VLM as a zero-shot reasoner to automatically generate preference pairs from a pretrained model's rollouts, which are then used to finetune the model via Direct Preference Optimization (DPO). We finetune our models on the Waymo Open End-to-End Driving Dataset (WOD-E2E) and evaluate performance against held-out human preference annotations using rater feedback score (RFS) and average displacement error (ADE). Our experiments confirm that the VLM's trajectory selection is a high-quality proxy for human preference. Our final model, VL-DPO, yields an 11.94% increase in RFS and a 10.01% reduction in ADE over the pretrained model.

Evaluating large language models (LLMs) on natural-language logical reasoning is essential because rule-governed tasks require conclusions to follow strictly from stated premises. Many existing logical-reasoning benchmarks are generated by templating natural-language items from sampled formulas, provide only coarse or unaudited formal annotations, and are now quickly saturated by frontier reasoning models. We present LLMEval-Logic, a Chinese logical reasoning benchmark built from realistic situational scenarios. Its pipeline forward-authors and expert-audits natural-language items together with their reference formalizations, verifies annotated answers with Z3, constructs expert rubrics for natural-to-formal grading, and hardens selected items through a closed-loop adversarial workflow. The benchmark is released in two paired subsets: a 246-item Base subset shipped with 1,400 expert-developed rubric atoms, and a 190-item Hard subset with 938 multi-step sub-questions over closed model spaces. Evaluating 14 frontier LLMs on LLMEval-Logic reveals substantial gaps in current models: the best model reaches only 37.5% Hard Item Accuracy, and even with reference symbols the highest joint Z3+Rubric formalization score among evaluated models reaches only 60.16%. Our benchmark is publicly available at https://github.com/llmeval/LLMEval-Logic.

Flash floods in Bangladesh's haor wetlands show up with almost no warning. They wreck the annual boro rice harvest. Current setups, built for riverine floods, miss backwater dynamics entirely. These basins are flat. Water does not behave like it does on the Brahmaputra. We built HaorFloodAlert, a deseasonalized machine learning ensemble that forecasts 72-hour flood probability for the Sunamganj Haor (approximately 8,000 km2). Temperature was acting as a seasonal cheat code - it inflated accuracy by 6.9 pp just because floods happen in warm months. We caught that. We also built an upstream Barak River Sentinel-1 SAR proxy from Silchar, Assam, giving about 36 hours of lead time. Otsu-thresholded SAR change detection validates at 84-91 percent spatial match. The operational ensemble (RF 0.5625 + XGBoost 0.4375) hits 89.6 percent LOOCV accuracy, 87.5 percent recall, and 0.943 AUC-ROC on 77 real Sentinel-1 events. A three-tier alert pipeline and a BRRI-calibrated boro rice damage estimator are included.

We revisit the renormalisation of models with two U(1) gauge symmetries, in a formulation with non-canonical gauge kinetic terms which is covariant under field reparametrisations among the two gauge bosons. This approach is convenient to study the appearance of kinetic mixing in scale evolution, because a coupling matrix is attributed to the gauge kinetic terms. We obtain simple MSbar renormalisation group equations up to two-loop, which can be solved to give effective millicharges at low energy which depend on the running couplings and heavy mass scales of the model. This formulation allows to construct ``invariants'' out of running Lagrangian parameters, which are invariant under generic gauge field reparametrisations, including rescalings, and which can be related directly to observables such as the millicharge.