探索全球前沿学术脉络

01.

arXiv (math.PR) 2026-06-16 DOI: arXiv:2606.16636

Free energy of non-convex multi-species spin glasses with centered Ising spins

作者:

Hong-Bin Chen↗Victor Issa↗Jean-Christophe Mourrat↗

arXiv:2606.16636v1 Announce Type: new Abstract: We identify the limit free energy of all multi-species spin glasses with centered $\pm 1$ spins. The result was previously known only under a convexity assumption on the covariance function of the Hamiltonian. We also obtain a one-species reduction of the formula for balanced multi-species models.

阅读与讨论 → 访问原文 →

02.

arXiv (CS.CL) 2026-06-25 DOI: arXiv:2606.25436

Evaluating Japanese Dialect Robustness Across Speech and Text-based Large Language Models

作者:

Tomoya Mizumoto↗Yusuke Fujita↗Hao Shi↗Lianbo Liu↗Atsushi Kojima↗Yui Sudo↗

Dialogue systems based on large language models (LLMs) have advanced significantly in recent years. However, dialectal variation remains a major challenge, particularly for systems that process spoken input. LLM-based speech language models (SLMs), which integrate LLMs with speech processing components, show promise for spoken language tasks, yet their ability to comprehend dialects has not been sufficiently studied. Moreover, it remains unclear how the dialectal understanding of the base LLM affects SLM performance. This study investigates the dialectal robustness of both LLMs and SLMs using Japanese dialects as a test case. We define robustness as the ratio of performance on dialectal versus standard inputs, enabling fair comparisons. Our experiments show that SLM robustness correlates with that of their text-based counterparts. Furthermore, training with dialectal data and fine-tuning the speech encoder each improves robustness in SLMs.

阅读与讨论 → 访问原文 →

03.

arXiv (CS.AI) 2026-06-15 DOI: arXiv:2606.13732

When Sample Selection Bias Precipitates Model Collapse

作者:

Xinbao Qiao↗Xianglong Du↗Wei Liu↗Jingqi Zhang↗Peihua Mai↗Meng Zhang↗Yan Pang↗

arXiv:2606.13732v1 Announce Type: new Abstract: The proliferation of recursive training on synthetic data can alleviate data scarcity but risks model collapse, where repeated training erodes distributional tails and homogenizes outputs. Data selection is widely viewed as a remedy, yet its reliability depends critically on the reference distribution used by the verifier. We show that in low-resource verification regimes, where each verifier observes only a small, fragmented, and biased slice of the target manifold, selection itself becomes biased. This situation naturally arises in low-resource data silos such as healthcare consortia or proprietary financial institutions, where raw data cannot be pooled and local references are inherently incomplete. As a result, selection preferentially retains samples aligned with the local manifold while pruning globally relevant tail modes, turning from a safeguard against collapse into a mechanism that precipitates it. We theoretically prove that such siloed selection accelerates collapse and induces power-law diversity decay. As an initial mitigation, we construct Wasserstein proxy references from multiple silos without sharing raw data. Empirical results confirm that local-reference selection fails on skewed distributions, whereas collaborative proxy references mitigate diversity degradation, suggesting that recursive synthetic-data pipelines require particular caution when real-data coverage is fragmented or scarce.

阅读与讨论 → 访问原文 →

04.

arXiv (quant-ph) 2026-06-17 DOI: arXiv:2606.17095

Variational Quantum Eigensolver-Based Quantum Bootstrap Embedding for Molecules

作者:

Derek Peng↗

arXiv:2606.17095v1 Announce Type: cross Abstract: Simulating strongly correlated molecular systems on near-term quantum hardware remains challenging due to modern hardware's limited quantum volume and moderate-fidelity qubits. One potential way to circumvent this challenge is through bootstrap embedding (BE). Bootstrap embedding breaks molecules into smaller fragments that are then embedded into the "bath" of other fragments in an iterative way. Bootstrap embedding is appealing for quantum simulation because fragmenting the system reduces the qubit requirements for any given fragment. In this work, we develop a quantum bootstrap embedding (QBE) workflow that uses variational quantum eigensolver (VQE) fragment solvers and study the algorithmic choices that determine the overall VQE-QBE algorithm's success. To improve efficiency, we introduce FastAdaptVQE, a sparse matrix-accelerated form of the adaptive variational quantum eigensolver (ADAPT-VQE) that replaces symbolic commutator evaluation with direct statevector linear algebra, and MatrixFreeAdaptVQE, a matrix-free extension that removes the sparse-matrix memory bottleneck that appears when treating larger fragments. We also modify the ADAPT-VQE operator selection step by replacing the purely greedy choice with a look-ahead strategy. Benchmarks on $H_4$ and $F_2$ reach chemical accuracy, within 1 kcal/mol of bootstrap embedding results using a full configuration interaction (FCI) solver. These results show that combining QBE with VQE can accurately calculate energies of molecular systems. This research lays the foundation for extending energy calculations to larger molecular systems and quantum materials on near-term quantum hardware.

阅读与讨论 → 访问原文 →

05.

arXiv (CS.AI) 2026-06-16 DOI: arXiv:2606.16826

ATOM-Bench: A Real-World Benchmark for Atomic Skills and Compositional Generalization in Manipulation Policies

作者:

Zenan Wu↗Bingqing Wei↗Lu Liu↗Zheqi He↗Xi Wang↗Jiakang Liu↗Zehui Li↗Guocai Yao↗Jing-Shu Zheng↗Xi Yang↗Yongtao Wang↗

arXiv:2606.16826v1 Announce Type: cross Abstract: Generalist manipulation policies are increasingly presented as foundation models for robotic control, but their real-world generalization remains difficult to diagnose. A policy may succeed on demonstrated tasks while still failing to execute fine-grained atomic skills or recombine learned skills in new task structures. We introduce ATOM-Bench, a real-world benchmark for evaluating both atomic skills and compositional generalization in manipulation policies. ATOM-Bench factorizes tabletop manipulation into motor atoms and instruction atoms, and contains 30 atomic tasks and 24 held-out compositional tasks across paired single-arm and dual-arm robot tracks. We collect 3,000 human demonstrations for atomic fine-tuning and release both the demonstration data and evaluation rollout data to support reproducible real-world evaluation. Policies are fine-tuned on atomic tasks and evaluated on both atomic skill acquisition and held-out compositional tasks. We further introduce Atomic Score (AS) and Compositional Failure Share (CFS) to distinguish failures caused by weak atomic skills from failures caused by limited compositional reuse. Through 2,700 physical rollouts on five representative manipulation policies, we find that current policies can acquire simple instruction-grounding skills, but still struggle with fine-grained motor atoms, counting, and logical filtering. More importantly, strong atomic performance does not reliably transfer to held-out compositional tasks. ATOM-Bench provides a diagnostic testbed for studying whether failures arise from weak motor execution, poor instruction grounding, or limited compositional reuse.

阅读与讨论 → 访问原文 →

06.

arXiv (CS.CL) 2026-06-12 DOI: arXiv:2606.10403

KCSAT-ML: Probing Reasoning Models with Nationwide-Cohort Human Difficulty

作者:

Sanghee Park↗Geewook Kim↗Kee-Eung Kim↗

Math reasoning benchmarks have proliferated, yet most lack a per-item difficulty signal grounded in actual human performance. We introduce KCSAT-ML, a decade (2014-2025) of Korean College Scholastic Ability Test (KCSAT; Suneung) mathematics: 664 problems with a 339-item core set carrying official per-item error rates from nationwide cohorts of hundreds of thousands of examinees. We pair the benchmark with Difficulty-aligned Reasoning Gain (DRG): a score-orthogonal metric that asks whether a model's mistakes concentrate on the items humans found hard, or on items humans found easy. Together they expose, across a wide range of VLMs (and LLMs via OCR), three patterns: (i) low-budget accuracy collapses on the high-human-error tail at every model size; (ii) test-time scaling (TTS) raises token use roughly linearly with cohort error rate, while accuracy gains follow a non-monotonic curve; (iii) within a single family, TTS flips between anti-scaling on the hardest items and overthinking on easier ones – two faces of the same alignment failure. On DRG, models with near-identical accuracy can sit at near-opposite values: one model gets wrong what humans also find hard, while another solves the hardest items yet fails on items humans find easy – a contrast that aggregate accuracy hides. Our code and dataset builder will be open-sourced at https://github.com/naver-ai/KCSAT-ML.

阅读与讨论 → 访问原文 →

07.

arXiv (CS.AI) 2026-06-18 DOI: arXiv:2605.17986

LivePI: More Realistic Benchmarking of Agents Against Indirect Prompt Injection

作者:

Lei Zhao↗Abhay Bhaskar↗Edgar Dobriban↗

arXiv:2605.17986v3 Announce Type: replace-cross Abstract: AI agents such as OpenClaw are increasingly deployed in local workflows with access to external tools. This creates indirect prompt-injection (IPI) risk: an agent may execute harmful instructions embedded in untrusted inputs such as email, downloaded files, webpages, repositories, or group-chat messages. Existing evaluations are often small, purely simulated, or focused on a narrow set of channels. We introduce LivePI (Live Prompt Injection), a structured benchmark for IPI risk in a production-like but test-controlled environment. LivePI covers seven input surfaces, twelve attack/rendering families, and five malicious goals, including protected-information exfiltration, unauthorized security-control changes, unsafe code retrieval or execution, inbox-summary exfiltration, and cryptocurrency transfer. We run LivePI on a real virtual machine with live but test-controlled email, chat, web, local-file, repository, and wallet interfaces. Across GPT-5.3-Codex, Claude Opus 4.6, Gemini 3.1 Pro, Kimi K2.5, and GLM-5, total attack success rates range from 10.7% to 29.6%. Group-chat injection is uniformly successful across the evaluated backbones in our deployment, and repository-link attacks produce high-severity failures despite a small denominator. We also evaluate a two-layer defense consisting of prompt-level filtering and pre-execution tool-call authorization. In the GPT-5.3-Codex setting, the defense intercepts all tested malicious-goal completions in LivePI before execution while preserving benign utility on PinchBench-derived workloads.

阅读与讨论 → 访问原文 →

08.

arXiv (CS.AI) 2026-06-24 DOI: arXiv:2606.24062

RAVEN: A Regime-Aware Variable-context Expert Network for Financial Time Series Forecasting

作者:

Cheng He↗Zhenyu Guan↗Xijie Liang↗Defu Lian↗Jiajia Li↗Enhong Chen↗Patrick P. C. Lee↗Geng Hu↗Zehao Chen↗

arXiv:2606.24062v1 Announce Type: cross Abstract: Financial time series forecasting presents structural challenges absent from standard benchmarks. Log-returns are non-stationary, exhibit exceptionally low signal-to-noise (SNR) ratios, and are governed by regime-dependent temporal dependencies. We identify a key limitation of state-of-the-art (SOTA) time series models in financial settings. A fixed context window is mismatched to the time-varying optimal look-back of non-stationary price processes. We propose the Regime-Aware Variable-context Expert Network (RAVEN), a Mixture-of-Experts framework designed to adaptively determine the temporal context for each input sample. Instead of relying on a fixed look-back horizon, RAVEN constructs a hierarchy of nested contiguous windows whose lengths are determined by the data itself. Specifically, RAVEN scores patches by learned importance in reverse chronological order and applies the Cumulative Importance Thresholding (CIT) mechanism to derive nested prefix windows, each routed to a scale-specialized expert. A Global Compressed Representation (GCR) branch runs in parallel over the full context, preserving global temporal coherence that local experts cannot guarantee. Because the nested routing induces structured overlap among expert inputs, we introduce a Correlation-Aware Weighting (CAW) to align variable-length expert outputs and penalize pairwise cosine similarity prior to aggregation. Experiments on cumulative log-return prediction (HS300, S&P500) and fund sales forecasting demonstrate that RAVEN achieves SOTA performances, improves Pearson correlation by 9.2% on HS300 and 20.2% on S&P500, and reduces MSE by 18.2% on fund sales forecasting, while achieving the best results in 14 of 16 metrics on four PEMS traffic benchmarks.

阅读与讨论 → 访问原文 →

09.

Nature Medicine 2026-06-12 DOI: HASH:fde650b09ce972374e2d68e7e4b14d0c

Efficacy and target engagement of dopamine agonist pramipexole for anhedonic depression: a randomized placebo-controlled trial

作者:

Filip Ventorp↗

Anhedonia is a core and disabling symptom of mood disorders with limited treatment options. We evaluated the efficacy and safety of the dopamine agonist pramipexole in patients with mood disorders characterized by clinically significant anhedonia. In this single-center, randomized, double-blind, placebo-controlled trial, adults with major depressive disorder, dysthymia or bipolar depression and elevated Snaith−Hamilton Pleasure Scale (SHAPS) scores were assigned (1:1) to flexible dose, once-daily oral pramipexole as add-on treatment or placebo for 9 weeks. The primary outcome was change in SHAPS score from baseline to week 9. Analyses were conducted in the modified intention-to-treat population. Eighty-five participants were randomized, and 82 were included in the analysis. The primary outcome was met: pramipexole was associated with a greater reduction in SHAPS scores compared to placebo (mean difference: −4.04, 95% confidence interval: −6.89 to −1.18, P = 0.006, Hedges’ g = 0.62). Exploratory analyses indicated that pramipexole was associated with increased light physical activity and relative preservation of reward-related ventral striatal activation. Improvements in anhedonia were sustained during a 6-month open-label extension. Pramipexole was generally well tolerated compared to placebo. Pramipexole significantly improved anhedonia and showed a favorable safety profile, supporting its potential as an augmentation strategy in mood disorders. ClinicalTrials.gov identifiers: NCT05355337 and NCT05825235 . Pramipexole, in patients with major depressive disorder, dysthymia or bipolar depression, reduced Snaith−Hamilton Pleasure Scale scores significantly compared to placebo.

阅读与讨论 → 访问原文 →

10.

arXiv (CS.CL) 2026-06-19 DOI: arXiv:2606.19831

Leverage Is Not Reach: A Control-Window Law for Single-Neuron Steering in Language Models

作者:

Hongliang Liu↗

Aligned language models gate behaviors such as refusal and language routing through sparse feed forward neurons, yet no theory predicts when a single neuron intervention controls a behavior coherently rather than collapsing the output. We develop a budget normalized control window framework for single neuron steering. A dose along one write direction reduces to one control coordinate: the alignment between the residual stream and the write, driven along a universal saturation curve in units of a coherence budget set by the residual norm divided by the write norm. Coherent control exists when a behavior trigger lies below the collapse ceiling. The same coordinate governs benign mode switches and refusal; the ceiling follows from weights and one generic forward pass, while triggers are measured at rollout. On fifteen held out neurons, the predicted ceiling has mean absolute error 0.14, about 0.07 in bulk layers, and the committed open or closed verdict holds on eleven against a ten of fifteen majority baseline. Closed cases expose three failure modes rather than violations: collapse before trigger, too little depth to propagate, or a normalization that caps how far one neuron can push. The law explains why local gradient attribution anti predicts control: true controllers write off the readout axis and carry a near zero first order gradient. A forward only contrastive screen made precise by the window recovers controllers that attribution misses. On refusal, the hardest case, intervention success is typed, not scalar: coherent bypass and strict actionable reach separate, so a neuron can flip refusal in fluent, on task text with no actionable content, and genuine actionable reach appears only for three of six audited Llama pivots and only at later rollout horizons. Single neuron steering is therefore a budgeted, typed audit of controllability rather than a fixed dose anecdote.

阅读与讨论 → 访问原文 →

11.

arXiv (CS.AI) 2026-06-25 DOI: arXiv:2606.24938

Privacy-preserving federated tensor decomposition of single-cell immune data: recovering multicellular programs across institutions

作者:

Axel Faes↗Stephanie M. van den Berg↗Maryam Amir Haeri↗

arXiv:2606.24938v1 Announce Type: cross Abstract: Tensor decomposition of donor $\times$ cell-type $\times$ gene single-cell data recovers multicellular programs: coordinated axes of inter-individual transcriptional variation that span cell types and stratify disease. Yet immune single-cell atlases are increasingly multi-institution, multi-ancestry, and governed, so patient cells often cannot be pooled. We present a federated estimator: each site computes a local program subspace, and a coordinator merges these by stacked SVD under federated global-mean centering, provably equivalent (up to truncation) to the centralised decomposition. This centering makes the merge robust to site-label confounding (program AUC $0.957$ vs.\ $0.861$ for naive per-site centering). Only program subspaces leave a site, and aggregation is compatible with secure aggregation. On a 261-donor systemic lupus erythematosus atlas it recovers the canonical interferon program (ISG enrichment AUC $0.998$; case–control separation $0.958$; bootstrap $\DeltaAUC=-0.000$, 95\% CI $[-0.004,+0.012]$ vs.\ centralised), across institution-scale and multi-ancestry partitions, and across three real COVID-19 sites (subspace correlation $0.989$). It recovers the program when no site observes all cell types (correlation $1.000$, exact by construction), which fixed-feature federated PCA cannot. On an interstitial-lung-disease atlas the recovered program predicts disease better than the best single cell type (AUC $0.96$ vs.\ $0.91$; gap 95\% CI excludes zero) and the advantage survives federation; a liver cohort is consistent ($p=0.005$). Membership-inference shows secure aggregation cuts attack AUC from $0.91$ to $0.61$. The method enables cross-institution, cross-ancestry recovery of multicellular immune programs without sharing cells.

阅读与讨论 → 访问原文 →

12.

arXiv (CS.AI) 2026-06-15 DOI: arXiv:2606.14078

Rethinking Backdoor Adversarial Unlearning through the Lens of Catastrophic Forgetting in Continual Learning

作者:

Zhenqian Zhu↗Yamin Hu↗Yujiang Liu↗Luping Wei↗Wenbo Hou↗Bin Li↗Haodong Li↗Wenjian Luo↗

arXiv:2606.14078v1 Announce Type: cross Abstract: Existing studies reveal that current backdoor defenses exhibit limited robustness and often fail against specific types of attacks. More concerningly, prevailing safety tuning strategies tend to provide only superficial safety protection, as they fall short of completely eliminating the backdoor effects. In this work, we present a novel formulation of backdoor learning and unlearning as a sequential, three-stage process from a continual learning perspective. Within this framework, we formally define complete backdoor unlearning and further derive the necessary conditions for achieving it based on the mechanism of catastrophic forgetting. Guided by these insights, we propose Blind Inversion-Backdoor Adversarial Unlearning (BI-BAU), which formulates the generation of adversarial examples satisfying the unlearning conditions as a blind inversion problem. We solve this by integrating the bi-level optimization process of adversarial training into an Expectation-Maximization (EM) algorithm framework to optimize the maximum a posteriori (MAP) objective. Furthermore, BI-BAU is extended to untargeted adversarial scenarios with unknown target classes, as well as to multi-modal contrastive learning tasks, enhancing its applicability to real-world deployment scenarios where pre-trained models may be compromised. Extensive experiments demonstrate that our method exhibits general applicability across a wide spectrum of backdoor attacks and can effectively and thoroughly eliminate the backdoor effects from a backdoor model.

阅读与讨论 → 访问原文 →

13.

arXiv (CS.CL) 2026-06-19 DOI: arXiv:2606.20065

Generative Engine Optimization at Scale: Measuring Brand Visibility Across AI Search Engines

作者:

Pratyush Kumar↗

People increasingly get answers straight from AI search engines like ChatGPT, Claude, Perplexity, and Gemini rather than scrolling search results. Brands that once focused on search engine optimization (SEO) must now optimize for how these engines represent, cite, and recommend them – a shift variously called Generative Engine Optimization (GEO), Answer Engine Optimization (AEO), and AI Search Visibility. We treat AEO and AI Visibility as part of GEO, and study how to measure brand visibility across AI engines: what they value when they cite a brand, which sources they rely on, and what content large language models surface. The hard case is everyone outside the already-authoritative top brands – SMEs, D2C brands, creators, and early-stage startups. We analyze 100K+ prompt responses across 100+ brands tracked on Ranqo between March and May 2026. First visibility runs form a clear three-tier brand-stature ladder: global household names (e.g., Stripe, Nike) appear in 73% of relevant AI answers on their first run; established mid-market and regional brands (e.g., Olipop, Klaviyo) in 44%; niche and small brands in just 11% – about 30 percentage points per step. When engines cite sources, about 78% go to corporate websites; among non-corporate sources YouTube leads, ahead of Reddit, editorial media, and Wikipedia. The highest-leverage page is the ranked "best-of" listicle, the most-cited content format at about 21% of all citations. Sentiment is the unstable signal: whether a brand is framed positively or negatively flips about 6.7 times more often than whether it is mentioned at all. These findings provide a first large-scale baseline for measuring GEO: AI brand visibility can be measured, differs by platform, and varies strongly by brand maturity. We close by proposing seven v1.1 protocols to test whether specific recommendations can causally improve AI visibility.

阅读与讨论 → 访问原文 →

14.

PLOS Computational Biology 2026-06-12 DOI: HASH:5c8c918a8a7100b969116a452c936bdc

Ten simple rules for executing an inherited research plan in computational biology

作者:

Sahar Javaheri Tehrani↗

by Sahar Javaheri Tehrani, Toni Ingolf Gossmann Trainees in computational biology frequently inherit research plans whose aims, datasets, analytical strategies, and technical constraints were defined before their arrival. These plans often emerge from grants, collaborations, legacy codebases, shared high-performance computing environments, or partially completed analyses. While such plans provide a useful scaffold, they rarely specify all implementation details, prior assumptions, evaluation criteria, or dependencies needed for reliable execution. The transition from inheriting a partially articulated plan to producing reproducible results therefore creates an execution gap: a phase in which trainees must reconstruct what the project is, which elements are fixed, which remain negotiable, and which technical or organizational assumptions need to be tested before full-scale analysis begins. In this Ten Simple Rules article, we provide a practice-oriented framework for stabilizing inherited computational biology projects before workflows, benchmarks, and decision paths become entrenched. We do not claim that the individual practices described here are novel in isolation. Rather, our contribution is to organize familiar practices into a sequenced framework for a recurrent but under-articulated phase of computational research: inherited-plan execution. Computational biology makes this phase especially important because projects often combine heterogeneous datasets, fragile software environments, undocumented preprocessing choices, benchmarking assumptions, distributed collaborators, and asymmetrical access to contextual knowledge. By making this transition visible and operational, the rules aim to help trainees, supervisors, and collaborators reduce ambiguity, test feasibility, document decisions, and support reproducible and equitable project execution under real-world constraints.

阅读与讨论 → 访问原文 →

15.

arXiv (CS.LG) 2026-06-12 DOI: arXiv:2606.13172

Detecting Explanatory Insufficiency in Learned Representations: A Framework for Representational Vigilance

作者:

Jacques Raynal↗Pierre Slangen↗Elsa Raynal↗Jacques Margerit↗

arXiv:2606.13172v1 Announce Type: new Abstract: Learned representations are central to modern machine learning and are commonly evaluated through predictive performance, robustness, uncertainty estimation, or generalization. However, a learned representation may remain operationally successful while progressively failing to organize persistent residual structures that are not fully captured by conventional evaluation metrics. This article introduces VER, the Vigilant Evaluator of Representations, a conceptual framework for monitoring representational adequacy in learned representations. VER does not propose a new learning algorithm, loss function, or model architecture. Instead, it formalizes a diagnostic process through which persistent residual structures may be identified, analyzed, and interpreted as potential indicators of explanatory insufficiency. The framework distinguishes representational inadequacy from ordinary prediction error, uncertainty, noise, and distribution shift. It introduces a monitoring sequence based on representation identification, explanatory-domain delimitation, residual-structure detection, explanatory-resistance evaluation, and vigilance signaling. VER is intended as a contribution to representation diagnostics in machine learning. Its objective is not to replace existing evaluation methods but to complement them by treating representational adequacy as an explicit object of inquiry. A path toward empirical evaluation through representational-vigilance benchmarks is also outlined.

阅读与讨论 → 访问原文 →

16.

arXiv (CS.AI) 2026-06-16 DOI: arXiv:2606.15549

CmdNeedle: Measuring the Incompleteness of Command Denylists for AI Agents

作者:

Chuyang Chen↗Zhiqiang Lin↗

arXiv:2606.15549v1 Announce Type: cross Abstract: The adoption of AI agents is increasing rapidly. Terminal AI agents, i.e., AI agents that run in terminal environments, are a widely used type of AI agents. Terminal AI agents rely heavily on shell command execution to interact with the host systems. They adopt a three-list command-gating mechanism to mitigate security risks introduced by command execution, with denylists serving as the load-bearing component. However, modern operating systems often ship a large, ever-expanding set of shell commands with complex functionalities. Our observation is that even a built-in denylist of Claude Code, well-maintained by its developers, can overlook bypass commands that invalidate its effectiveness. Such negligence leads to fragile command denylists that cannot even block operations that practitioners expect them to block. This paper presents the first systematic characterization of command denylist fragility in terminal AI agents. The paper formalizes the command denylist fragility problem and proposes an LLM-driven pipeline, CmdNeedle, to detect such fragility. It prompts the LLM to propose possible bypasses and iteratively repairs them using feedback from a validator that executes them in a sandbox. In the evaluation, we applied CmdNeedle to 1,709 real-world command denylists (containing 13,332 denylist rules) collected from GitHub. The evaluation shows several key findings, including that 69.0–98.6% of the denylists are fragile, that this fragility occurs consistently across projects and agents, and the validity of several possible root causes for this fragility. Our pipeline and findings will hopefully facilitate future research and practice regarding the command denylists used by AI agents.

阅读与讨论 → 访问原文 →

17.

arXiv (math.PR) 2026-06-17 DOI: arXiv:2606.17517

Killed resolvents and measure-valued stopping gains for reflected optimal stopping with max-type rewards

作者:

Louis Shuo Wang↗Jiguang Yu↗Ye Liang↗

arXiv:2606.17517v1 Announce Type: new Abstract: We study an infinite-horizon optimal stopping problem for a normally reflected two-dimensional diffusion in the positive quadrant with nonsmooth max-type reward \(G(x_1,x_2)=x_1\vee \alpha x_2\). The paper develops a conditional measure-theoretic framework for the associated reflected obstacle problem. The main innovation is to show that the stopping gain \(\Gamma=c+rG-\mathcal LG\) is a signed measure, not a function: the kink of \(G\) generates an explicit negative surface measure on \(\Delta=\{x_1=\alpha x_2\}\). We then prove that the correct potential representation uses the resolvent of the reflected diffusion killed on first entry into the stopping set, rather than the unrestricted reflected resolvent. Under explicit monotonicity, regularity, and measure-superharmonicity assumptions, we derive an epigraph representation, a continuation-side boundary-trace condition, and a candidate verification theorem. The framework clarifies hidden regularity and uniqueness assumptions in multidimensional nonsmooth optimal stopping.

阅读与讨论 → 访问原文 →

18.

arXiv (CS.AI) 2026-06-16 DOI: arXiv:2606.16286

FlowMPC: Improving Flow Matching policies with World Models

作者:

Chandon Hamel↗

arXiv:2606.16286v1 Announce Type: cross Abstract: Flow Matching (FM) is a powerful approach for behavior cloning in multimodal action spaces [Jiang et al., 2025], but because it is not trained to directly maximize expected return, there is still room to improve how FM policies act at test time. This work investigates whether a learned world model can improve FM policies by enabling Model Predictive Path Integral (MPPI) planning over candidate action sequences proposed by the policy. Building on TD-MPC2 [Hansen et al., 2024], I introduce FlowMPC, a framework that combines an imitation-learned FM policy with a learned world model for test-time planning in ManiSkill manipulation tasks [Tao et al., 2025]. Across PickCube and PickSingleYCB, adding the world model improved performance over the FM policy alone, with especially clear gains in end-of-episode success. These results suggest that world-model-based planning can effectively complement flow-based imitation policies without modifying the FM training objective.

阅读与讨论 → 访问原文 →

19.

arXiv (CS.AI) 2026-06-16 DOI: arXiv:2606.16730

Attention is Just Another Name for Coupling?: A Fast-Slow ODE Perspective on Hierarchical Pretraining

作者:

Zhengyuan Gao↗

arXiv:2606.16730v1 Announce Type: cross Abstract: Causal self-attention is a coupling mechanism: each token's hidden state is updated by a learned mixture of preceding tokens at the same timescale. This paper asks whether a second, temporally slower coupling-a slow sub-system operating on a temporally-downsampled view of the sequence and fed back into the fast path through a zero-initialised gate-complements it. The question is framed in the language of singularly perturbed ordinary differential equations (ODEs), where the fast variable $x$ evolves at the token rate, the slow variable $y$ evolves at one update per $P$ tokens, and the timescale ratio $\varepsilon = 1/P$ is enforced structurally by causal block-mean pooling. The paper instantiates the fast-slow ODE formalism as a concrete neural network: a fast path of standard causal attention over $T$ tokens, a slow path of full attention over $T/P$ pooled tokens ($P^2 \times$ cheaper per layer), and a zero-initialised additive gate. In addition, under a linear-generator assumption on the fast dynamics, we prove that the equilibrium manifold $x = \phi(y)$ is exactly the master-equation (ME) stationary distribution $p_{\mathrm{st}}(y)$; in that regime a learned MLP $\phi_\theta(y)$ is a variational approximation of it (the trained block is not a generator, so this identity is the structured limit, not a claim about the network as trained). Empirically, at $500$k tokens the coupling is neutral – the gate stays closed and the coupled and frozen ablations are within run-to-run noise – at a wall-clock cost comparable to a dense baseline. The contribution is the precise, gap-marked mapping itself, not a performance gain.

阅读与讨论 → 访问原文 →

20.

Nature (Science) 2026-06-15 DOI: HASH:b404890b82e6b94261153672622dea2e

Nanocrystal-tailored recombination for all-perovskite tandem solar modules

作者:

Ke Xiao↗

The commercialization of all-perovskite tandem solar modules is hindered by the reliance on the conventional gold-based tunnel recombination junction (TRJ)1,2. Specifically, this TRJ introduces substantial near-infrared parasitic absorption3 and suffers from interfacial instability4, limiting both photocurrent generation and operational durability. Here, we develop a solution-processed interconnecting layer based on surface-engineered indium oxide (In2O3) nanocrystals featuring high optical transparency, wherein controlled nanocrystal morphology and tailored ligand chemistry enable smooth interfacial contact and favorable energy level alignment. Critically, we introduce a phosphonic acid additive into the lead–tin (Pb–Sn) perovskite precursor, which synergistically improves the electronic contact with the In2O3 recombination layer, thereby enhancing hole extraction. In addition, the additive regulates perovskite crystallization to mitigate residual strain during film formation, ensuring high-quality large-area deposits. This coordinated interfacial and crystallization engineering strategy simultaneously enhances carrier recombination efficiency at the interconnection layer, improves carrier extraction, and promotes large-area film uniformity in all-perovskite tandems. As a result, a 65-cm2 all-perovskite tandem solar module achieves a certified power conversion efficiency of 26.2%5, with an open-circuit voltage of 2.182 V, a fill factor of 77.4%, and a short-circuit current density of 15.6 mA cm-2 in terms of averaged subcell performance, measured by Japan Electrical Safety and Environment Technology Laboratories (JET). This marks a significant advance toward scalable perovskite tandem photovoltaics.

阅读与讨论 → 访问原文 →

21.

arXiv (CS.CV) 2026-06-25 DOI: arXiv:2606.25478

TACO: Towards Task-Consistent Open-Vocabulary Adaptation in Video Recognition

作者:

Minghao Zhu↗Xiao Lin↗Mengxian Hu↗Xun Zhou↗Liuyi Wang↗Xiaoyan Qi↗Chengju Liu↗Qijun Chen↗

Adapting CLIP for open-vocabulary video recognition necessitates a delicate balance between newly acquired video knowledge and the pretrained generalization. While existing studies pursue this generalization-specialization trade-off with additional regularizations or constraints, we argue that they overlook the deviation of representations beyond the fine-tuning data distribution, resulting in suboptimal adaptation effects. We believe such deviation is inherited from the inconsistency between the fine-tuning and evaluation objectives, where model optimization is restricted to the known training distribution but evaluated on unseen ones. In this paper, we introduce TACO, a simple yet effective framework to mitigate the potential negative effects induced by this inconsistency. Our key insight is that adaptation should preserve OOD-relevant alignment beyond the training distribution. To this end, we propose Relative Structure Distillation, which regularizes the relative geometry of the representation space and suppresses harmful alignment shift during training. We further decouple the representation space from the optimization space with a lightweight specialization projection, allowing task-specific adaptation without directly overspecializing the representations used at test time. TACO establishes state-of-the-art performance on diverse benchmarks under cross-dataset and base-to-novel settings. Code will be released at https://github.com/ZMHH-H/TACO.

阅读与讨论 → 访问原文 →

22.

arXiv (CS.CV) 2026-06-18 DOI: arXiv:2606.19195

Moebius: 0.2B Lightweight Image Inpainting Framework with 10B-Level Performance

作者:

Kangsheng Duan↗Ziyang Xu↗Wenyu Liu↗Xiaohu Ruan↗Xiaoxin Chen↗Xinggang Wang↗

While 10B-level industrial foundation models have pushed the boundaries of image inpainting, their prohibitive computational costs severely hinder practical deployment. Constructing a highly optimized task-specific specialist offers a promising solution; however, extreme structural compression inevitably triggers a severe representation bottleneck. To conquer this, we propose Moebius, a highly efficient lightweight inpainting framework. We systematically reconstruct the diffusion backbone by introducing the Local-$\lambda$ Mix Interaction ($L\lambda MI$) block. Comprising Local-$\lambda$ and Interactive-$\lambda$ modules, it elegantly summarizes spatial contexts and global semantic priors into fixed-size linear matrices, preserving complex latent interactions while drastically shedding parameters. Furthermore, to unlock the full representational capacity of this highly compact architecture, we synergistically pair it with an adaptive multi-granularity distillation strategy. Operating strictly within the latent space to avoid expensive pixel-space decoding, this strategy dynamically balances multiple gradient-based losses to achieve high-fidelity alignment. Extensive experiments across natural and portrait benchmarks demonstrate that this optimal synergy enables Moebius to rival or even surpass the generation quality of the 10B-level industrial generalist FLUX.1-Fill-Dev. Remarkably, Moebius achieves this using less than 2\% of the parameters (0.22B vs. 11.9B) while delivering a $>15\times$ acceleration in total inference time, setting a new efficiency standard for high-fidelity inpainting. Project page at https://hustvl.github.io/Moebius.

阅读与讨论 → 访问原文 →

23.

arXiv (CS.CL) 2026-06-12 DOI: arXiv:2606.12576

Helping Figures Tell their Story! Paper-Grounded Video Generation Explaining Complex Scientific Figures

作者:

Ishani Mondal↗Javad Baghirov↗Jordan Boyd-Graber↗

Scientific figures compress complex pipelines into a single canvas, yet understanding them requires paper-grounded, step-by-step narration aligned with visual highlights a capability missing from current video generation systems and benchmarks. To address this, we introduce paper-grounded figure-to-video generation: generating narrated, region-grounded walkthrough videos from a figure and its paper. We propose MINARD (Multimodal Interpretation of Narrated Architecture via Region Decomposition), a pipeline that generates paper-grounded narrations and sequentially grounds them to figure regions. We also release FigTalk, a benchmark with new sequential and component-level grounding metrics derived. On FigTalk, MINARD generates humanlike, paper-faithful narrations and outperforms narration-conditioned figure spatial grounding compared to existing approaches in both automatic and human evaluation

阅读与讨论 → 访问原文 →

24.

arXiv (CS.LG) 2026-06-15 DOI: arXiv:2606.13799

The Program Is Still There: A Conservation Law for Program Discovery

作者:

Jorge Miguel Silva↗

arXiv:2606.13799v1 Announce Type: cross Abstract: Finding the shortest program that generates a sequence is uncomputable, and for six decades that fact has been mistaken for a wall around finding any generating program. It is not a wall but a price, and this paper measures it. For every algorithm that learns about a candidate program only through its score, a class spanning Levin search, evolutionary methods, simulated annealing, and the cross-entropy method, we define the coupling width of a search problem and prove an unconditional worst-case lower bound, exponential in that width with base one less than the domain size. From it follows a conservation law: structural knowledge injected into a search trades one for one against the search it removes, and their sum can never fall below the length of the program sought. Levin's 1973 upper bound and the lower bound proved here are the two ends of one conserved quantity, closing on each other as the instruction set grows. The only escape is to read a candidate's structure rather than its score, and its price, which we prove for generic targets, is incompleteness. A deterministic engine built on this theory recovers a generating program, certified by compressing its data and predicting an unseen continuation, for 2,383 of 3,914 sequences across four independent populations, including 244 of the 256 elementary cellular automata, with measured discovery cost rising along program length more than an order of magnitude inside the score-oracle worst case.

阅读与讨论 → 访问原文 →

25.

arXiv (CS.CL) 2026-06-25 DOI: arXiv:2606.25524

Cliff Tokens: Identifying Single-Token Failure Triggers in LLM Mathematical Reasoning

作者:

Jaeyong Ko↗Pilsung Kang↗Yukyung Lee↗

Large language models (LLMs) reach high accuracy in mathematical reasoning, but individual traces on the same problem diverge; some arrive at the correct answer while others fail. Prior work analyzes failure at the step, chunk, or sentence level, or at tokens where failure has already occurred. Neither identifies the precise token that triggers the shift toward failure. We introduce the cliff token, a token where the token-wise potential drops significantly under an adaptive threshold that scales with the local token-wise potential, based on a one-sided two-proportion z-test. Across seven models and three mathematical reasoning benchmarks (GSM1K, MATH500, AIME 2025), cliff tokens act as failure triggers; deleting the first cliff token and resampling recovers pass@64 to 1.0, while keeping it limits recovery to between 0.71 and 1.00. We further introduce a cliff taxonomy of deterministic, uncertain, and sampled-off cliffs, defined by greedy choice and token entropy. Each type has distinct probabilistic characteristics, and the taxonomy generalizes across model scales. Finally, we validate the taxonomy via single-token preference optimization at cliff positions (Cliff-DPO). Trained on GSM8K, Cliff-DPO improves accuracy across benchmarks by up to +6.6. Optimizing at uncertain and sampled-off cliffs improves reasoning, while deterministic cliffs do not.

阅读与讨论 → 访问原文 →