探索全球前沿学术脉络

01.

arXiv (quant-ph) 2026-06-11 DOI: arXiv:2606.12045

A Geometric Family of Correlations Containing the Quantum Singlet

作者:

E. Aldo Arroyo↗

arXiv:2606.12045v1 Announce Type: new Abstract: We introduce a geometrically constrained hidden-variable framework that generates a family of correlations parametrized by a boundary function, within which the quantum singlet correlation appears as a particular member. Exact expressions for the correlation function are derived. Several structural results are established, including admissibility conditions, symmetry properties, a universal stationary point of the associated CHSH function, and an exact relation between the CHSH value at $\nu=\pi/4$ and a geometric contrast measure defined on the underlying hidden-variable distributions. Rather than treating the quantum singlet correlation as an isolated target to be reproduced, the present framework places it within a broader geometric structure of correlations. These results suggest the existence of a nontrivial geometric structure underlying the family of correlations and motivate the search for a principle capable of selecting the quantum singlet solution from within that family.

阅读与讨论 → 访问原文 →

02.

arXiv (math.PR) 2026-06-25 DOI: arXiv:2606.25204

Exponential Rank Bounds for Random Matrices

作者:

Achintya Raya Polavarapu↗

arXiv:2606.25204v1 Announce Type: new Abstract: Fix $b\in(0,1)$, let $1\leq k\leq n$, and let $A=(A_{ij})$ be an $n\times n$ random matrix with independent real entries satisfying $$ \sup_{x\in\mathbb{R}}\mathbb{P}\{A_{ij}=x\}\leq b0$ such that $$ \mathbb{P}\{\operatorname{rank} A\leq n-k\}\leq \exp(-cnk), \qquad 1\leq k\leq n. $$

阅读与讨论 → 访问原文 →

03.

medRxiv (Medicine) 2026-06-22 DOI: HASH:92dad982b75d8c867dfd3de7268f145e

The Unsteady Return of Command-Following: Recovery and Instability of Bedside Motor Command-Following After Acute Brain Injury

作者:

Gorenshtein↗Adiniaev↗Omar↗Barash↗Klang↗Daniel↗

Background/Objective: Following a verbal command marks the bedside transition from unresponsiveness to overt recovery of consciousness after acute brain injury. Its timing across phenotypes, stability once present, and dependence on sedation are uncharacterized at scale. Methods: Retrospective cohort of adults with acute brain injury, first intensive care unit stay, MIMIC-IV. Command-following was the Glasgow Coma Scale motor response "Obeys Commands." Among patients not following commands at admission, cumulative incidence was estimated with death or hospice and discharge without recovery as competing events. Instability was quantified as transient first recovery and threshold crossings; examinations were tagged for concurrent sedation. Principal findings were externally validated in the multicenter eICU Collaborative Research Database. Results: Of 13,900 brain-injured patients with three or more motor examinations, 5,498 (39.6%) were not following commands at admission. The cumulative incidence of first command-following was 43.5% by 24 hours and 65.0% by 14 days, ranging at 14 days from 36.9% in anoxic injury to 77.2% in ischemic stroke (anoxic versus ischemic stroke at 72 hours, difference 0.41; adjusted P = .002). Among 3,573 patients who recovered, the first recovery was transient in 22.2%, and 62.4% crossed the threshold repeatedly. Non-following was strongly associated with sedation, consistent with an arousal-dependent examination. In eICU, the 14-day incidence was 64.8%, and transient first recovery was 22.7%, closely matching the primary cohort. Conclusions: After acute brain injury, overt bedside command-following returns early but unsteadily, with phenotype-dependent timing, threshold fluctuation, and strong dependence on sedation. A single charted observation is an unreliable index of the underlying state.

阅读与讨论 → 访问原文 →

04.

arXiv (CS.CL) 2026-06-16 DOI: arXiv:2606.16576

Can LLM Agents Infer World Models? Evidence from Agentic Automata Learning

作者:

Reef Menaged↗Gili Lior↗Shauli Ravfogel↗Roee Aharoni↗Gabriel Stanovsky↗

We propose agentic automata learning to evaluate the extent to which tool-calling LLM agents can uncover hidden environments through interaction. In our setup, an agent should uncover a hidden deterministic finite automaton (DFA) by interacting with an oracle through (1) membership queries ("Does this string belong to the target language?") and (2) equivalence queries ("Is this the target DFA?"). This yields a scalable testbed with controlled task complexity, measurable interaction efficiency, and strong baselines (classic automata-learning algorithms). Evaluating state-of-the-art LLMs, we find that performance drops sharply as DFA size increases. Reasoning models are markedly stronger than non-reasoning models, yet trajectory analyses reveal recurring failures in query planning, evidence integration, and hypothesis construction. Overall, our results show that current LLM agents can sometimes perform non-trivial interactive discovery, but remain far less robust and efficient than classic algorithms for the task.

阅读与讨论 → 访问原文 →

05.

arXiv (CS.AI) 2026-06-11 DOI: arXiv:2604.13733

Vision-Language-Action Jump-Starting for Reinforcement Learning Robotic Agents

作者:

Angelo Moroncelli↗Roberto Zanetti↗Marco Maccarini↗Loris Roveda↗

arXiv:2604.13733v2 Announce Type: replace-cross Abstract: Reinforcement learning (RL) enables high-frequency, closed-loop control for robotic manipulation, but scaling to long-horizon tasks with sparse or imperfect rewards remains difficult due to inefficient exploration and poor credit assignment. Vision-Language-Action (VLA) models leverage large-scale multimodal pretraining to provide generalist, task-level reasoning, but current limitations hinder their direct use in fast and precise manipulation. In this paper, we propose Vision-Language-Action Jump-Starting (VLAJS), a method that bridges sparse VLA guidance with on-policy RL to improve exploration and learning efficiency. VLAJS treats VLAs as transient sources of high-level action suggestions that bias early exploration and improve credit assignment, while preserving the high-frequency, state-based control of RL. Our approach augments Proximal Policy Optimization (PPO) with a directional action-consistency regularization that softly aligns the RL agent's actions with VLA guidance during early training, without enforcing strict imitation, requiring demonstrations, or relying on continuous teacher queries. VLA guidance is applied sparsely and annealed over time, allowing the agent to adapt online and ultimately surpass the guiding policy. We evaluate VLAJS on six challenging manipulation tasks: lifting, pick-and-place, peg reorientation, peg insertion, poking, and pushing in simulation, and validate a subset on a real Franka Panda robot. VLAJS consistently outperforms PPO and distillation-style baselines in sample efficiency, reducing required environment interactions by over 50% in several tasks. Real-world experiments demonstrate zero-shot sim-to-real transfer and robust execution under clutter, object variation, and external perturbations.

阅读与讨论 → 访问原文 →

06.

arXiv (CS.LG) 2026-06-16 DOI: arXiv:2606.14900

GRASP: Gradient-Aligned Sequential Parameter Transfer for Memory-Efficient Multi-Source Learning

作者:

Mary Isabelle Wisell↗Nicholas Jacobs↗Aayush Manandhar↗Salimeh Yasaei Sekeh↗

arXiv:2606.14900v1 Announce Type: new Abstract: Multi-source transfer learning faces a fundamental scalability bottleneck: existing approaches require either loading all K source models into memory simultaneously during parameter fusion, requiring O(K) memory, or deploying all models at inference time, making production deployment infeasible. We propose GRASP (Gradient-Aligned Sequential Parameter Transfer), which achieves superior knowledge integration while maintaining O(1) memory consumption through three key innovations: (1) sequential processing that merges one source at a time into an evolving target model, (2) parameter-wise gradient alignment that selectively transfers only parameters whose optimization directions align with the target domain, avoiding negative transfer, and (3) iterative fine-tuning that adapts transferred knowledge before integrating the next source. Extensive experiments across three continual learning benchmarks (Yearbook, CLEAR-10, CLEAR-100) spanning 10 to 108-year temporal distribution shifts and four architectures (1.3M to 25.6M parameters) demonstrate that GRASP achieves 93.5% mean accuracy over all datasets and architectures compared to ensemble method's 71.7% accuracy while requiring only constant memory versus K models for standard multi-source fusion. Critically, GRASP's sequential previously merged models and scales to arbitrarily many sources without memory growth, making it uniquely suitable for resource-constrained deployment and continually evolving source domains.

阅读与讨论 → 访问原文 →

07.

arXiv (CS.CL) 2026-06-24 DOI: arXiv:2606.24523

Poster: Exploring the Limits of Audio-Based Detection of Turkish Phone Call Scams

作者:

Arda Eren↗Micheal Cheung↗Youqian Zhang↗Grace Ngai↗Eugene Yujun Fu↗

Scam phone calls exploit vulnerable communities worldwide, yet research on detection has focused almost exclusively on English and other high-resource languages. In low-resource settings such as Turkish, detection is especially difficult, as annotated data is scarce and technological defenses remain limited. This research investigates how large language models (LLMs) can support scam detection in Turkish by introducing the first public multi-modal dataset of 100 aligned audio-transcript pairs of scam and benign conversations. We evaluate seven LLMs spanning three model families: Gemini 2.5 (Flash, Flash-Lite, Pro), GPT-4o, and Qwen (Max, Plus, Turbo), under three input conditions: raw audio, automatic speech-to-text transcripts, and transcripts refined by a native speaker. Our results suggest that transcript-based inputs consistently outperform direct audio processing, while human-corrected and uncorrected transcripts perform comparably. By centering a low-resource language and real world threat, this work highlights the urgent need for culturally and linguistically inclusive AI safety research and more robust multi-modal systems for fraud prevention.

阅读与讨论 → 访问原文 →

08.

arXiv (CS.AI) 2026-06-17 DOI: arXiv:2606.17409

Discrete Autoregressive Transformer for Generative Mechanism Synthesis

作者:

Anar Nurizada↗Anurag Purwar↗

arXiv:2606.17409v1 Announce Type: cross Abstract: Planar path synthesis requires mechanisms whose coupler curves match a prescribed trajectory; the mapping from curve to linkage is inherently one-to-many across four-, six-, and eight-bar topologies. We address this design problem with simulation-grounded evaluation on a curated corpus of over one million mechanisms, reporting Chamfer distance and dynamic time warping after forward kinematics and geometric alignment. We formulate synthesis as conditional autoregressive sequence modeling: joint coordinates are uniformly quantized to tokens and generated by a decoder-only transformer with a variational-autoencoder (VAE) latent of the target curve and an explicit mechanism-type token. Training combines token cross-entropy with a Gaussian-smoothed bin auxiliary loss that respects ordinal structure among bins. At inference, a bounded latent-noise schedule decodes all mechanism types at each noise level; we retain the top five candidates by geometric error, yielding diverse accurate families without dataset lookup. On held-out tests, aggregate mean Chamfer distance is $0.0132$ and mean dynamic time warping is $0.153$; a latent $k$-nearest-neighbor baseline that conditions on training-set neighbor latents in VAE space achieves matched-topology mean Chamfer distance $0.0071$ and mean dynamic time warping $0.117$ using the same decoder.

阅读与讨论 → 访问原文 →

09.

arXiv (quant-ph) 2026-06-15 DOI: arXiv:2606.14088

OQMD: Single-Qubit Rotation Control Improves Low-CNOT Multiclass Quantum Classification

作者:

Michael A. Magid↗Melissa Zeynep Ertem↗Jun Suzuki↗

arXiv:2606.14088v1 Announce Type: new Abstract: Near-term variational classifiers incur substantial error and latency from two-qubit gates, yet practitioners often assume that additional entangling depth is the default route to higher accuracy. This work studies Optimal Quantum Measurement Decoding (OQMD): optimizing how quantum outcomes are mapped to classical labels by training a readout layer before measurement, jointly with the variational circuit, without adding CNOTs. Experiments use trainable triple single-qubit rotations as one concrete, hardware-native realization of OQMD; other single-qubit parametrizations fit the same classical outer loop. On the Iris benchmark with a 30-point stratified test split, the best observed 0-CNOT configuration with OQMD reaches 83.33\% accuracy, with a 96\% at 9 CNOTs, exceeding the best 18-CNOT controls (56.67\%) and the best 18-CNOT configuration with OQMD (66.67\%) under a common protocol. A six-point CNOT-depth series from 0 to 18 (fixed optimizer, iteration budget, random-seed count, and ZXZ readout) shows that the highest raw scores need not occur at the largest template, so aggregate complexity is not summarized by CNOT count alone. Because run-level accuracies are discrete and non-Gaussian, we emphasize best-observed scores and, where a global comparison of pooled runs is required, Mann–Whitney $U$ tests rather than parametric tests on means. Across architectures, OQMD shows statistically consistent but magnitude-dependent gains: large peak lifts on minimal circuits coexist with a small pooled mean shift on complex 18-CNOT runs ($p\approx 0.03$) that is not ``universal'' in the sense of uniformly large practical effects.%

阅读与讨论 → 访问原文 →

10.

arXiv (CS.LG) 2026-06-18 DOI: arXiv:2606.18514

N(CO)$^2$: Neural Combinatorial Optimization with Chance Constraints to Solve Stochastic Orienteering

作者:

Anas Saeed↗Marcos Abel Zuzu\'arregui↗Stefano Carpin↗

arXiv:2606.18514v1 Announce Type: cross Abstract: Neural combinatorial optimization (NCO) offers a promising alternative to traditional heuristic-based methods for solving complex graph optimization problems by proposing to learn heuristics through data. This class of problems frequently arises in automation, as it can be used to model a variety of applications. While NCO has been extensively studied for deterministic combinatorial optimization problems, there are only a few works that aim to solve stochastic combinatorial optimization problems. In this work, we present N(CO)$^2$: Neural Combinatorial Optimization with Chance cOnstraints to solve the Stochastic Orienteering Problem (SOP) without the use of hand-crafted heuristics. By integrating a reinforcement learning (RL) framework, the model optimizes path selection under uncertainty, effectively balancing exploration and exploitation. Empirical results demonstrate that our method generalizes well across diverse SOP instances, achieving competitive performance compared to the state-of-the-art mixed-integer linear program (MILP) for the task. The proposed approach reduces human effort in heuristic design while enabling adaptive and efficient decision-making in uncertain environments.

阅读与讨论 → 访问原文 →

11.

arXiv (CS.LG) 2026-06-16 DOI: arXiv:2510.14217

Spectral Analysis of Molecular Features: When Richer Features Do Not Guarantee Better Generalization

作者:

Asma Jamali↗Tin Sum Cheng↗Rodrigo A. Vargas-Hern\'andez↗

arXiv:2510.14217v2 Announce Type: replace Abstract: The spectral properties of feature embeddings offer critical insights into model generalization and representation quality. While deep learning models are widely used for molecular property prediction, kernel methods remain competitive in low-data regimes, yet their spectral behavior is largely unexplored. We present the first comprehensive spectral analysis of kernel ridge regression across diverse representations-including molecular fingerprints (ECFP), pretrained transformers, graph neural networks, and 3D descriptors-evaluated on QM9 and 3 MoleculeNet benchmarks. Surprisingly, richer spectral features do not consistently yield better generalization performance, contradicting common representation heuristics used in self-supervised learning (SSL). Across 4 spectral metrics, only ECFP-based kernels show a strictly positive correlation with performance. Transformer and global 3D representations exhibit mixed behavior, whereas local 3D representations show consistently negative correlations. Truncation analysis further emphasizes this disparity: for local 3D representations on thermodynamic targets, fewer than 2\% of eigenvalues (and occasionally as few as 0.02\%) are needed to recover 95\% of performance, whereas ECFP and transformer kernels require significantly more. By demonstrating a strong dependence on both task and representation, our results challenge the heuristic that richer spectra inherently improve generalization, providing new guidance for evaluating representations in SSL and in label-limited scientific tasks.

阅读与讨论 → 访问原文 →

12.

arXiv (CS.CV) 2026-06-24 DOI: arXiv:2606.04764

Do Foundation Models See Biology? Evaluating Attention Coherence with Spatial Transcriptomics in Glioblastoma

作者:

Dilakshan Srikanthan↗Amoon Jamzad↗Paul Wilson↗Nooshin Maghsoodi↗Robert Policelli↗Gabor Fichtinger↗John F. Rudan↗Parvin Mousavi↗

Whether attention maps from pathology foundation models capture genuine biology remains unknown, yet this question is critical for clinical trust and regulatory approval. We propose a spatial transcriptomics-based framework for orthogonal, hypothesis-free evaluation of attention and apply it to five pathology foundation models (CONCH v1.5, UNI v2, Virchow2, GigaPath, H-Optimus-1) and a ResNet50 baseline. Using attention-based multiple instance learning, we train single-task and multi-task models to predict five molecular alterations in glioblastoma on the CPTAC cohort, validate on an independent TCGA cohort, and evaluate biological coherence of attention maps against 87 transcriptional signatures using co-registered Visium spatial transcriptomics data from 18 samples. Internally, no single encoder dominates across all tasks, and external validation inverts internal performance rankings. Attention maps show a five-fold enrichment gradient from pathways (Cohen's d=0.329) to individual genes (d=0.055), indicating that attention captures emergent multi-gene transcriptional programs rather than individual molecular events. Spatially smooth attention maps do not imply biological coherence, and different encoders attend to distinct biological compartments. Our framework provides objective, quantitative assessment of what foundation models learn from histopathology, moving the field beyond qualitative saliency map review.

阅读与讨论 → 访问原文 →

13.

arXiv (CS.CL) 2026-06-24 DOI: arXiv:2606.23695

Quantifying Prior Dominance in RAG Systems

作者:

Barak Or↗

Retrieval-Augmented Generation (RAG) grounds Large Language Models in external knowledge, yet current evaluations rely on discrete heuristics that suffer from ''epistemic blindness'' - failing to distinguish genuine contextual information extraction from parametric memory recall. To address this, we introduce the Normalized Context Utilization (NCU) metric, leveraging continuous token log-probabilities across zero-shot, oracle, and adversarial conditions to strictly quantify contextual information gain. Evaluating architectures ranging from 1.5B to 72B parameters alongside a proprietary commercial API reveals that for strict factual extraction (without Chain-of-Thought reasoning), traditional scaling laws exhibit extreme diminishing returns: highly efficient Small Language Models (SLMs) match or outperform high-capacity architectures. Furthermore, we demonstrate that ``Prior Dominance'' correlates with model scale and proprietary alignments. The evaluated commercial API not only overrode explicit external evidence in nearly half of adversarial conflicts, but also frequently suffered from systemic confidence collapse (Negative Transfer) when its parametric priors were contradicted. Our findings highlight the structural epistemic advantage and superior contextual adherence of SLMs in strict extraction workflows.

阅读与讨论 → 访问原文 →

14.

arXiv (CS.CL) 2026-06-24 DOI: arXiv:2511.21056

Bilevel Data Curation for LLM Fine-tuning: Offline Selection and Online Self-Refining Generation

作者:

Quan Xiao↗Yutong Xuan↗Gaowen Liu↗Ramana Rao Kompella↗Tianyi Chen↗

Supervised fine-tuning (SFT) datasets are critical to the downstream performance of large language models, yet they often contain low-quality or harmful question-response pairs. To improve SFT data quality, we develop a unified bilevel framework that combines offline data selection with the online self-refining generation. In the offline setting, bilevel data selection (BDS) selects question-response pairs from the offline SFT dataset to maximize the validation performance. We theoretically show that the optimal model given by BDS outperforms direct data mixing approach in useful data coverage. Moreover, we provide a global convergence analysis for gradient-based BDS approach for one-layer Transformer, showing that the epsilon-global optimum of offline BDS is achievable in finite time. Although efficient, offline BDS discards potentially harmful questions together with responses, thereby reducing question diversity. We address this limitation by refining the responses to selected questions using online self-refining generation framework. However, BDS is inefficient to update the response weights when responses are regenerated online. To address this issue, we introduce bilevel multi-objective optimization (BMO) for response-level weighting. We show that BMO recovers the same validation-aligned solution as BDS, but admits a closed-form importance-ratio weight that adapts to regenerated responses. Experiments on LLM quality enhancement and safety-aware fine-tuning demonstrate that the proposed framework consistently improves both data quality and downstream fine-tuning performance.

阅读与讨论 → 访问原文 →

15.

arXiv (CS.CV) 2026-06-19 DOI: arXiv:2606.19495

LooseControlVideo: Directorial Video Control using Spatial Blocking

作者:

Shariq Farooq Bhat↗Niloy J. Mitra↗Kalyan Sunkavalli↗

Precise 3D spatial orchestration in text-to-video generation remains a significant challenge, particularly for multi-object scenes where semantic layout and temporal dynamics are often entangled. While existing depth-conditioned models achieve good structural fidelity, they necessitate dense, frame-accurate guidance that is labor-intensive to author for dynamic events involving deformable objects. We present LooseControlVideo, a framework that enables intuitive and expressive control by using sparse, oriented 3D boxes as a "blocking" proxy. This allows users to author high-level layout and trajectory while leveraging a video generative model to generate realistic occlusions, dynamics and interactions. We achieve this by fine-tuning a Wan 2.2 backbone on a video dataset annotated with DNOCS, a novel encoding for 3D size, orientation and depth-ordered occlusions. Furthermore, our method allows for localized refinement, such as adjusting a jump trajectory or adding an interaction, with minimal disruption to the global scene context. Extensive evaluations on the nuScenes, HO-3D, and BEHAVE benchmarks demonstrate that LooseControlVideo significantly outperforms existing 2D-box and flow-based baselines. Our findings indicate a 1.2x to 3x improvement in Trajectory Error; 2x improvement in Rigid Motion Consistency; and a 1.5x to 2x increase in Occlusion Accuracy over current state-of-the-art layout-conditioned models, demonstrating that oriented 3D primitives provide good geometric prior for complex, multi-agent video authoring.

阅读与讨论 → 访问原文 →

16.

arXiv (math.PR) 2026-06-16 DOI: arXiv:2504.07255

A Tail-Respecting Splitting Numerical Scheme for Lévy-Driven SDEs With Superlinear Drifts

作者:

Olga Aryasova↗Oleksii Kulyk↗Ilya Pavlyukevich↗

arXiv:2504.07255v3 Announce Type: replace Abstract: We present an explicit numerical approximation scheme, denoted by $\{X^n\}$, for the effective simulation of solutions $X$ to a multivariate stochastic differential equation (SDE) with a superlinearly growing $\kappa$-dissipative drift, where $\kappa>1$, driven by a multiplicative heavy-tailed Lévy process that has a finite $p$-th moment, with $p>0$. We show that the strong $L^{p_X}$-convergence $\sup_{t\in[0,T]}\mathbf E \|X^n_t-X_t\|^{p_X}=\mathcal O (h_n^{\gamma})$ holds for any $p_X\in (0,p+\kappa-1)$, which is exactly the range where the $p_X$-moment of the solution is known to be finite. Additionally, for any $p_X\in (0,p)$ we establish strong uniform convergence: $\mathbf E\sup_{t\in[0,T]} \|X^n_t-X_t\|^{p_X}=\mathcal{O} ( h_n^{\delta} )$. In both cases we determine the convergence rates $\gamma$ and $\delta$. In the special case of SDEs driven solely by a Brownian motion, our numerical scheme preserves super-exponential moments of the solution. The scheme $\{X^n\}$ is realized as a combination of a well-known Euler method with a Lie-Trotter type splitting technique.

阅读与讨论 → 访问原文 →

17.

arXiv (math.PR) 2026-06-12 DOI: arXiv:2601.00793

Voronoi Percolation: Topological Stability and Giant Cycles

作者:

Benjamin Schweinhart↗Morgan Shuman↗

arXiv:2601.00793v2 Announce Type: replace Abstract: We study the topological stability of Voronoi percolation in higher dimensions. We show that slightly increasing p allows a discretization that preserves increasing topological properties with high probability. This strengthens a theorem of Bollobás and Riordan and generalizes it to higher dimensions. As a consequence, we prove a sharp phase transition for the emergence of i-dimensional giant cycles in Voronoi percolation on the 2i-dimensional torus.

阅读与讨论 → 访问原文 →

18.

arXiv (CS.LG) 2026-06-16 DOI: arXiv:2602.00781

Fast Non-Episodic Finite-Horizon RL with K-Step Lookahead Thresholding

作者:

Jiamin Xu↗Kyra Gan↗

arXiv:2602.00781v2 Announce Type: replace Abstract: Online reinforcement learning in non-episodic, finite-horizon MDPs remains underexplored and is challenged by the need to estimate returns to a fixed terminal time. Existing infinite-horizon methods, which often rely on discounted contraction, do not naturally account for this fixed-horizon structure. We introduce a modified Q-function: rather than targeting the full-horizon, we learn a K-step lookahead Q-function that truncates planning to the next K steps. To further improve sample efficiency, we introduce a thresholding mechanism: actions are selected only when their estimated K-step lookahead value exceeds a time-varying threshold. We provide an efficient tabular learning algorithm for this novel objective, proving it achieves fast finite-sample convergence: it achieves minimax optimal constant regret for $K=1$ and $\mathcal{O}(\max((K-1),C_{K-1})\sqrt{SAT\log(T)})$ regret for any $K \geq 2$. We numerically evaluate the performance of our algorithm under the objective of maximizing reward. Our implementation adaptively increases K over time, balancing lookahead depth against estimation variance. Empirical results demonstrate superior cumulative rewards over state-of-the-art tabular RL methods across synthetic MDPs and RL environments: JumpRiverswim, FrozenLake and AnyTrading. Code is provided on \href{https://github.com/jamie01713/K-Step-Lookahead}{github}.

阅读与讨论 → 访问原文 →

19.

arXiv (CS.AI) 2026-06-25 DOI: arXiv:2604.07945

Incremental Residual Reinforcement Learning Toward Real-World Learning for Social Navigation

作者:

Haruto Nagahisa↗Kohei Matsumoto↗Yuki Tomita↗Yuki Hyodo↗Ryo Kurazume↗

arXiv:2604.07945v2 Announce Type: replace-cross Abstract: As the demand for mobile robots continues to increase, social navigation has emerged as a critical task, driving active research into deep reinforcement learning (RL) approaches. However, because pedestrian dynamics and social conventions vary widely across different regions, simulations cannot easily encompass all possible real-world scenarios. Real-world RL, in which agents learn while operating directly in physical environments, presents a promising solution to this issue. Nevertheless, this approach faces significant challenges, particularly regarding constrained computational resources on edge devices and learning efficiency. In this study, we propose incremental residual RL (IRRL). This method integrates incremental learning, which is a lightweight process that operates without a replay buffer or batch updates, with residual RL, which enhances learning efficiency by training only on the residuals relative to a base policy. Through the simulation experiments, we demonstrated that, despite lacking a replay buffer, IRRL achieved performance comparable to those of conventional replay buffer-based methods and outperformed existing incremental learning approaches. Furthermore, the real-world experiments confirmed that IRRL can enable robots to effectively adapt to previously unseen environments through the real-world learning.

阅读与讨论 → 访问原文 →

20.

arXiv (CS.LG) 2026-06-19 DOI: arXiv:2606.19410

The Representational Limit of Scalar Interactions: An Interventional Decomposition

作者:

Potito Aghilar↗Sabino Roccotelli↗Stanislao Fidanza↗Vito Walter Anelli↗Sebastiano Stramaglia↗Tommaso Di Noia↗

arXiv:2606.19410v1 Announce Type: cross Abstract: Signed pairwise interaction scores fundamentally conflate uniqueness (U), redundancy (R), and synergy (S). We prove this on a minimal 3-way XOR structural causal model: faithful indices such as Shapley-Taylor return zero per pair, whereas projective indices such as Shapley Interaction spread the third-order effect into pair scalars that conflate the three mechanisms. We introduce Stochastic Hi-Fi, a post-hoc, retraining-free predictability decomposition that estimates per-feature U/R/S profiles by interventional masked inference. The estimator provides exact interventional semantics, finite-sample Monte Carlo bounds, strict variance reduction from coupled diamond sampling, and uniform finite-vocabulary convergence. Across tabular SCMs, Stochastic Hi-Fi recovers structure missed by scalar baselines (up to 411x larger interaction-magnitude recovery ratios). It also separates redundant and synergistic heads in the GPT-2 IOI circuit. On NIH ChestX-ray14, Stochastic Hi-Fi matches GradCAM on Pointing Game and improves substantially on Deletion AUC.

阅读与讨论 → 访问原文 →

21.

medRxiv (Medicine) 2026-06-20 DOI: HASH:7f3c9537ec4645216595bc52a1e9dcd2

EpiLink: a simulation-based compatibility model for genomic transmission clustering in infectious disease surveillance

作者:

Arthur↗Banks↗C. J↗Kao↗R. R↗

Identifying recently linked infections from pathogen genome sequences is central to infectious disease surveillance, yet many clustering approaches rely on fixed genetic distance thresholds whose relationship to transmission is often unclear. This limitation is especially important in rapidly growing outbreaks and superspreading events, where many cases may be sampled close together in time and share little genetic variation, making true transmission links difficult to distinguish from other closely related infections. Supervised models can improve discrimination, but they require labelled transmission data that are rarely available during outbreak response. We developed EpiLink, a threshold-free method that estimates whether two cases are compatible with recent transmission. Here, compatibility means how well the observed genetic distance and sampling-time difference between two cases fit what would be expected if they were linked by defined recent transmission scenarios. EpiLink simulates plausible recent transmission histories while accounting for uncertainty in infection timing, testing delay, and mutation accumulation, then assigns higher scores to pairs whose observed differences are typical of those simulations. EpiLink was evaluated using both synthetic and empirical SARS-CoV-2 outbreak data from the 2020 Boston epidemic. Two EpiLink variants were compared to a logistic regression model trained on labelled transmission data. One EpiLink variant assumed deterministic mutation accumulation, with genetic differences proportional to elapsed evolutionary time; the other accounted for stochasticity by sampling mutation counts from a Poisson distribution. The logistic regression model performed better at distinguishing linked from unlinked pairs, but EpiLink achieved comparable clustering accuracy. In the Boston data, EpiLink recovered clusters enriched for documented conference and skilled nursing facility outbreaks. EpiLink thus provides an interpretable, simulation-based approach for identifying recent transmission clusters when fixed thresholds are difficult to justify and labelled transmission data are unavailable.

阅读与讨论 → 访问原文 →

22.

arXiv (CS.AI) 2026-06-11 DOI: arXiv:2510.03520

Certifiable Safe RLHF: Semantic Grounding and Fixed Penalty Constraint Optimization for Safer LLM Alignment

作者:

Kartik Pandit↗Sourav Ganguly↗Arnesh Banerjee↗Shaahin Angizi↗Arnob Ghosh↗

arXiv:2510.03520v2 Announce Type: replace-cross Abstract: Ensuring safety is a foundational requirement for large language models (LLMs). Achieving an appropriate balance between enhancing the utility of model outputs and mitigating their potential for harm is a complex and persistent challenge. Contemporary approaches frequently formalize this problem within the framework of Constrained Markov Decision Processes (CMDPs) and employ established CMDP optimization techniques. However, these methods exhibit two notable limitations. First, their reliance on reward and cost functions renders performance highly sensitive to the underlying scoring mechanism, which must capture semantic meaning rather than being triggered by superficial keywords. Second, CMDP-based training entails tuning dual-variable, a process that is both computationally expensive and does not provide any provable safety guarantee for a fixed dual variable that can be exploitable through adversarial jailbreaks. To overcome these limitations, we introduce Certifiable Safe-RLHF (CS-RLHF) that introduces a cost model trained on a large-scale corpus to assign semantically grounded safety scores. In contrast to the lagrangian-based approach, CS-RLHF adopts a rectified penalty-based formulation. This design draws on the theory of exact penalty functions in constrained optimization, wherein constraint satisfaction is enforced directly through a suitably chosen penalty term. With an appropriately scaled penalty, feasibility of the safety constraints can be guaranteed at the optimizer, eliminating the need for dual-variable updates. Empirical evaluation demonstrates that CS-RLHF outperforms state-of-the-art LLM model responses rendering at-least 5 times efficient against nominal and jail-breaking prompts

阅读与讨论 → 访问原文 →

23.

bioRxiv (Bioinfo) 2026-06-11 DOI: HASH:f34a6c3c9d1b52daa265c99de17cd612

TifBERT: a self-supervised foundation model for normalization-robust bulk RNA-seq representation learning

作者:

Hosseini↗Sharma↗

Bulk RNA sequencing remains central to translational genomics, yet foundation-model development has largely focused on single-cell data. Existing transformer approaches for bulk RNA-seq often rely on expression discretization, numerical reconstruction, external gene embeddings, or restricted gene sets, limiting robustness across normalization schemes and cohorts. Here, we introduce TifBERT, a self-supervised framework for full-transcriptome bulk RNA-seq representation learning. TifBERT converts each unordered expression profile into a sample-specific gene sequence using term frequency-inverse document frequency (TF-IDF) ordering, prioritizing genes that are both highly expressed within a sample and selectively expressed across the cohort. It is then pretrained using masked gene modeling, predicting gene identities from transcriptomic context rather than reconstructing expression values. Pretrained on harmonized TCGA Pan-Cancer data spanning five RNA-seq normalization schemes, TifBERT learns contextual representations across approximately 10,000 genes without expression binning, landmark-gene restriction, or external biological embeddings. Across 33 TCGA cancer types, TifBERT achieved 90.83% accuracy, 0.996 macro AUC-ROC, and 0.903 MCC. It also captured pathway-level biology, achieving mean sample-wise and pathway-wise Pearson correlations of 0.754 and 0.762 across 1,387 PARADIGM pathway activities. Independent evaluation on GTEx healthy tissues showed preservation of tissue-level transcriptomic structure without retraining. In comparison with existing models, TifBERT achieves competitive subtype discrimination with substantially greater stability and produces markedly richer embedding geometry (effective rank 95.6 versus 6.3), without requiring expression discretization or in-distribution pretraining exposure. Together, TifBERT provides a scalable, normalization-independent foundation model for reusable bulk transcriptomic representation learning

阅读与讨论 → 访问原文 →

24.

arXiv (CS.CV) 2026-06-25 DOI: arXiv:2606.25545

TensorLDM: A Component-Wise Latent Diffusion Model for Volumetric DTI Reconstruction from Sparse DWIs

作者:

Junhyeok Lee↗Kyu Sung Choi↗

Reconstructing diffusion tensors from sparse DWIs is critical for accelerating Diffusion Tensor Imaging (DTI) in clinical settings, yet current deep learning approaches frequently yield anatomically inconsistent or physically implausible tensors. We introduce TensorLDM, a component-wise latent diffusion model that processes the six tensor components through two group-specific encoders (for diagonal and off-diagonal elements) while maintaining anatomical consistency via shared DWI conditioning. TensorLDM uses an Anatomy-Conditioned Autoencoder that encourages the latent to focus on tensor properties rather than re-encoding structural information. A shared Cross-Component Attention (CCA) mechanism, applied in both autoencoder refinement and diffusion fine-tuning, models inter-component dependencies, while a Mixture-of-Experts (MoE) DWI conditioner provides component-adaptive conditioning. On the Human Connectome Project (HCP) dataset under a single-shell, four-volume sparse acquisition, TensorLDM produces the most accurate downstream tractography and tensors with near-ground-truth physical validity (SPD-violation rate 1.54% vs. 1.40%), with the best or comparable voxel-wise reconstruction accuracy. Geodesic tensor error measured by the Log-Euclidean Metric (LEM) corroborates these gains.

阅读与讨论 → 访问原文 →

25.

arXiv (CS.CV) 2026-06-12 DOI: arXiv:2507.22028

From Seeing to Experiencing: Scaling Navigation Foundation Models with Reinforcement Learning

作者:

Honglin He↗Yukai Ma↗Brad Squicciarini↗Wayne Wu↗Bolei Zhou↗

Navigation foundation models trained on massive web-scale data enable agents to generalize across diverse environments and embodiments. However, these models, which are trained solely on offline data, often lack the capacity to reason about the consequences of their actions or adapt through counterfactual understanding. They thus face significant limitations in real-world urban navigation, where interactive and safe behaviors, such as avoiding obstacles and moving pedestrians, are critical. To tackle these challenges, we introduce the Seeing-to-Experiencing (S2E) learning framework to scale the capability of navigation foundation models with reinforcement learning. S2E combines the strengths of pretraining on offline videos and post-training through reinforcement learning. It maintains the model's generalizability acquired from large-scale real-world videos while enhancing its interactivity through reinforcement learning in simulation environments. Specifically, we introduce two innovations: (1) an Anchor-Guided Distribution Matching strategy for offline pretraining, which stabilizes learning and models diverse motion patterns through anchor-based supervision; and (2) a Residual-Attention Module for reinforcement learning, which obtains reactive behaviors from simulation environments without erasing the model's pretrained knowledge. Moreover, we establish a comprehensive end-to-end evaluation benchmark, NavBench-GS, built on photorealistic 3D Gaussian Splatting reconstructions of real-world scenes that incorporate physical interactions. It can systematically assess the generalizability and safety of navigation foundation models.

阅读与讨论 → 访问原文 →