Academic Intelligence · Curated Daily

探索全球前沿学术脉络

AcademicHub 汇聚顶级期刊与预印本平台的实时文献。定制您的专属科研雷达,利用大语言模型自动生成交叉领域文献分析简报。

01.
PLOS Medicine 2026-05-27

Sequential chemo-immunotherapy followed by standard versus reduced thoracic radiotherapy for older and/or frail stage III non-small-cell lung cancer: A randomized open-label cohort trial

作者:

by Wei-Xiang Qi, Shuyan Li, Mengdi Wang, Huan Li, Feifei Xu, Lei Yao, Biao Yu, Linlin Chen, Gang Cai, Cheng Xu, Xianwen Sun, Zhiyao Bao, Jiayi Chen, Yi Xiang, Shengguang Zhao Background The appropriateness of concurrent chemoradiotherapy (cCRT) for older or clinically vulnerable stage III unresectable non-small-cell lung cancer (NSCLC) patients remains contentious. Furthermore, the survival implications of de-escalating thoracic radiotherapy (RT) intensity in this population have not been conclusively elucidated. Methods and findings We conducted a phase II randomized, open-label, two-cohort (non-comparative) trial at a tertiary hospital in China (NCT05557552). Between September 30, 2022 and April 30, 2024, we enrolled 56 older and/or frail patients with stage III NSCLC who were ineligible for cCRT. The primary endpoint was the 1-year progression-free survival (PFS) rate estimated using the Kaplan–Meier method. Secondary endpoints included objective response rate (ORR), overall survival (OS), and safety. In the intention-to-treat (ITT) set, which included all 56 randomized patients who received at least one dose of study treatment, the 1-year PFS was 84.3% (95% confidence interval [CI] [70.3%, 98.3%]) in the standard RT group and 70.7% (95% CI [54.3%, 87.1%]) in the reduced RT group. In the per-protocol set (53 patients), the 1-year PFS was 82.9% (95% CI [68.9%, 98.8%]) in the standard RT group and 73.4% (95% CI [58.3%, 92.4%]), with a median follow-up of 24 months. Among 56 patients in the safety analysis set, 71.4% of patients experienced grade 3/4 adverse events (AEs) in the standard RT group and 53.6% in the reduced RT group. One patient (3.6%) in the reduced RT and three patients (10.7%) in the standardized RT experienced grade 5 AEs. The main limitations are the non-comparative design, small sample size, and lack of power to establish non-inferiority or superiority. Conclusion The current study suggested that reduced RT combined with sequential chemo-immunotherapy might be feasible for older/frail patients intolerant to cCRT, showing numerically similar survival outcomes. These exploratory findings warrant confirmation in larger, adequately powered randomized trials. Trial registration The trial had been registered on ClinicalTrials.gov on Sep 30, 2022.ClinicalTrials.gov NCT05557552

02.
arXiv (CS.AI) 2026-06-11

Ambient Diffusion Policy: Imitation Learning from Suboptimal Data in Robotics

arXiv:2606.12365v1 Announce Type: cross Abstract: We propose Ambient Diffusion Policy, a simple and principled method for imitation learning from suboptimal data in robotics. High-quality, task-specific robot data is expensive and time-consuming to collect, while suboptimal datasets with lower-quality or out-of-distribution demonstrations are abundant. Existing methods that co-train on both data sources in robotics often fail to separate the meaningful and the harmful features in the suboptimal samples. In contrast, our method extracts only the useful features by introducing a new axis to co-training in robotics: noise-dependent data usage. Ambient Diffusion Policy restricts the contribution of suboptimal data during training to only the high and low diffusion times. To rigorously justify our approach, we first observe that robot action data exhibits a spectral power law. This induces two important properties on the optimal Diffusion Policy that we exploit: a global-to-local hierarchy and locality. We theoretically formalize this discussion using a simplified model. Our experiments validate Ambient Diffusion Policy on four types of suboptimal action data (noisy trajectories, sim-to-real gap, task mismatch, and large-scale data mixtures) across six tasks. The results show that it effectively learns from arbitrary sources of suboptimal data. Notably, it outperforms existing co-training baselines by up to 33% when scaled to Open X-Embodiment - a large dataset with heterogeneous data quality and unstructured distribution shifts. Overall, Ambient Diffusion Policy increases the utility of suboptimal demonstrations and expands the set of usable data sources in robotics.

03.
arXiv (math.PR) 2026-06-19

Theory of uncertain probability: can we derive the probability density function of uncertain random experiments with continuously changing conditions?

作者:

arXiv:2606.20169v1 Announce Type: new Abstract: This paper aims to explore the formation mechanism of probability distribution in situations where the differences among random experiments are distinguishable, and these differences continue to evolve along with the dynamic changes in conditions and their mechanisms of action. To this end, we are motivated to devise a new theoretical system – theory of uncertain probability (TUP) with Kolmogorov's system and nonlinear theories as special cases. TUP develops a novel model that integrates probability and uncertainty as well as the known and unknown to more accurately depict numerous typical random phenomena under more realistic assumptions, and thus provides appropriate tools for greater variety of real needs. It also allows for pioneering interpretation of the causal mechanisms underlying many important distributional characteristics and incorporation of pathwise property to distribution model.

04.
arXiv (CS.CL) 2026-06-11

AI Coding Agents in Social Science: Methodologically Diverse, Empirically Consistent, Interpretively Vulnerable

The deployment of LLM-based agents in scientific analysis raises opposing concerns: that agents may reduce methodological diversity, or that they may amplify the analytic flexibility through which researchers reach motivated conclusions. We argue these worries target two empirically separable layers: a design layer of methodological choices, and a verdict layer in which a decision rule maps estimates to a substantive claim. We test both by running 20 independent executions of Claude Code and Codex on a prominent immigration and social-policy against a many-analysts human baseline. At the design layer, Codex matches human methodological diversity and Claude Code produces nearly three times as many specifications; both agents' effect estimates remain broadly aligned with the human consensus, and no agent model exactly matches any human model. A prompt-induced anti-immigration researcher prior reorganizes each agent's methodological decisions but, unlike for biased human analysts in the same data, does not shift aggregate estimates or final verdicts; nor do agents reroute along the methodological axes humans use to bias their estimates. At the verdict layer, an explicit confirmatory prompt flips Claude Code's verdicts from 10% to 90% support while leaving its coefficient distribution essentially unchanged, operating through rule omission rather than rule softening. AI agents can rival or exceed human methodological diversity at the design layer while remaining vulnerable at the verdict layer. In our setting, the locus of AI bias is not estimation but interpretation.

05.
arXiv (CS.AI) 2026-06-16

Z-Plane Neural Networks: Bounded Geometric Activation Replaces ReLU and LayerNorm

arXiv:2606.15669v1 Announce Type: cross Abstract: Modern deep neural networks rely on Euclidean scalar activations (e.g., ReLU) and global normalization techniques (e.g., LayerNorm) to prevent gradient instability in deep architectures. However, these mechanisms inherently cause dead neurons, discard critical directional information, and destroy the orthogonality of feature representations. Inspired by the frequency-modulation transmission of biological axons, we propose the Z-Plane Neural Network, which maps hidden states into 2D phasor bundles on a hypersphere. We introduce a novel geometric activation function, Radial Bounding($\mathbf{x} / \max(1, \|\mathbf{x}\|_2)$), which limits the energy magnitude while preserving the phase (direction). We demonstrate mathematically that this isotropic activation maintains 1-Lipschitz continuity and prevents gradient vanishing by preserving tangential gradients. Empirically, a 100-layer Z-Plane Multi-Layer Perceptron (MLP)-entirely devoid of ReLU and LayerNorm-successfully converges on the MNIST dataset with 98.34% accuracy and absolute numerical stability, proving that bounded geometric activation alone is sufficient for stable deep learning.

06.
arXiv (CS.CL) 2026-06-16

Cloze: An Open Research Platform for Studying Human-AI Conversations in Mental Health Contexts

Cloze is an open-source web platform for conducting controlled, monitored studies of human-AI conversation in mental health research contexts. Consumer large language model (LLM) products such as ChatGPT, Claude, and Gemini are built for individual productivity, and offer researchers little experimental control, inconsistent data export, and no shared safety scaffolding that holds across providers. Cloze gives research teams a single environment in which they configure which models participants converse with, how the AI is instructed, how conversations are scheduled over time, and which safety constraints apply unconditionally, while every message is captured with full provenance (model version, prompt configuration, timing). The platform currently supports OpenAI, Anthropic, Google, and locally hosted open-weight models served through Ollama behind a unified interface, and runs in the cloud or fully on premises so that participant data need never leave an institution. Cloze is research infrastructure for building an evidence base on human-AI interaction in mental health contexts. It is not a therapeutic product.

07.
bioRxiv (Bioinfo) 2026-06-19

ContinuumCellAgent: A Framework-Guided Agent for Long-Horizon Scientific Research

AI-scientist systems are beginning to automate parts of scientific research. We present ContinuumCellAgent, an autonomous agent that executes literature review, hypothesis formation, computational experimentation, manuscript drafting, and adversarial peer review as a single unattended run. Existing AI scientist systems remain difficult to diagnose because they lack modularity, systematic prompt grounding, and observability into long-running behavior. ContinuumCellAgent addresses these gaps with a modular supernode architecture for stage-wise backend swapping, protocols grounded in curated research-method checklists that also define reviewer rubrics, and a diagnostics layer that records file-based artifacts, message traces, and state transitions. We evaluate the system on open-domain QA benchmarks and biomedical/longevity case studies, showing that it can produce checkable research artifacts while exposing pipeline dynamics for rigorous AI co-scientist research.

08.
arXiv (CS.AI) 2026-06-16

A Causal Model of Theory of Mind in Conflict for Artificial Intelligence

arXiv:2606.16944v1 Announce Type: new Abstract: Theory of mind (ToM), the capacity to ascribe mental states to others and use those ascriptions for prediction and inference, is widely assumed to be essential for effective human-machine integration. Existing AI-ToM models address how to mentalize, but leave the question of when largely unaddressed. The central question is: under what situational and agent-level conditions is ToM engagement causally warranted in conflict? This paper presents a structural causal model formalized as a directed acyclic graph (DAG), treating ToM as a mechanism activated by situational and agent-level conditions rather than as an always-on capacity. The model specifies four exogenous variables capturing situational and agent-level conditions, five endogenous mediators, and a mechanistic ToM node producing engagement states through three distinct causal pathways: a tractability pathway, a reasoning-depth pathway, and an enabling-cause pathway. The primary outcome is epistemic accuracy, which decouples social reasoning from behavioral policy and generalizes across social phenomena beyond conflict. The framework gives AI systems a principled, resource-rational decision procedure for mentalizing, with implications for efficiency, trust, and the development of robust artificial social intelligence. Simulation validation, empirical human-machine teaming studies, and ethical considerations arising from conflict-optimized mentalizing are discussed.

09.
arXiv (CS.LG) 2026-06-11

Attention by Synchronization in Coupled Oscillator Networks

arXiv:2606.12059v1 Announce Type: new Abstract: We address transformer attention on energy-constrained physical substrates. Softmax attention requires exponentiation and global reduction, operations with high energy cost on von Neumann hardware and no natural physical analog. We show that Kuramoto synchronization dynamics (which arise in electrical, mechanical, superconducting, and charge-density-wave oscillator arrays, among other physical systems) implement a well-defined attention operation without either. The resulting mechanism, fixed-query oscillator attention, replaces softmax's arithmetic with the equilibration of a gradient flow on the sphere: queries are learned anchors fixed on the sphere, and free oscillators evolve under Kuramoto-Lohe dynamics until they settle at positions encoding attention weights via cosine similarity. Because the computation is equilibration, it requires no exponentiation; the only global operation is an affine normalization at readout. The fixed point is provably unique and globally attractive from almost every initial condition, a guarantee that holds across every physical realization. Empirically, at the minimal hardware configuration (oscillator dimension $d_{\mathrm{osc}}$ = 2), oscillator attention outperforms softmax on keyword spotting (+1.00 pp) and on subject-verb agreement (+5.27 pp on hard sentences, with zero training failures versus one in five for softmax). On causal language modeling, where softmax retains an advantage, oscillator attention closes the gap as $d_{\mathrm{osc}}$ grows: from +11.09 PPL at $d_{\mathrm{osc}}$ = 2 to +2.98 PPL at $d_{\mathrm{osc}}$ = 32 on WikiText-2, and from +2.39 PPL at $d_{\mathrm{osc}}$ = 2 to +0.57 PPL at $d_{\mathrm{osc}}$ = 32 on TinyStories. The main objective of this work is not to replace softmax in software but to provide a mathematically grounded blueprint for accurate attention on physical substrates.

10.
arXiv (CS.AI) 2026-06-11

Robust Instruction Compliance in Cooperative Multi-Agent Reinforcement Learning

arXiv:2605.12655v3 Announce Type: replace Abstract: Multi-agent reinforcement learning (MARL) in real-world use cases may need to adapt to external natural language instructions that interrupt ongoing behavior and conflict with long-horizon objectives. However, conditioning rewards on instructions introduces a fundamental failure mode as Bellman updates couple value estimates across instruction contexts, leading to inconsistent values when instructions interrupt macro-actions. We propose Macro-Action Value Correction for Instruction Compliance (MAVIC), which corrects Bellman backups at instruction boundaries by correcting the incoming instruction objective and restoring the continuation value under the current objective. Unlike reward shaping, MAVIC modifies the bootstrapping target itself, enabling consistent value estimation under stochastic instruction switching within a unified policy. We provide theoretical analysis and an actor-critic implementation, and show that MAVIC achieves high instruction compliance while preserving base task performance in increasingly complex cooperative multi-agent environments.

11.
arXiv (CS.AI) 2026-06-15

Patcher: Post-Hoc Patching of Backdoored Large Language Models

arXiv:2606.02995v2 Announce Type: replace-cross Abstract: Large language models remain vulnerable to jailbreak backdoor attacks, where adversaries poison safety alignment data to embed hidden triggers that bypass safety mechanisms. Existing defenses often require comprehensive attack information or multiple triggered examples, making them impractical when defenders only observe a single reported failure case without knowing whether it stems from a backdoor attack or a natural alignment bug. This paper presents Patcher, a post-hoc defense framework that repairs backdoored language models using only a single reported failure case and the model parameters. Patcher operates in two stages. First, it localizes backdoor triggers by computing response-conditioned gradient-based saliency scores and applying adaptive clustering to separate triggers from benign context. Second, it patches the model through a constrained fine-tuning objective that breaks the trigger-response association while preserving benign-task utility and robustness to non-triggered jailbreak attacks through KL-divergence constraints. We conduct extensive evaluations across multiple backdoor attack strategies and demonstrate that Patcher successfully localizes triggers and neutralizes backdoors while maintaining model utility. We further show robustness against adaptive attacks designed to evade our defense. This work represents a significant step toward practical defenses against training-time attacks in deployed language models.

12.
PLOS Medicine 2026-06-23

Parental body mass index and offspring childhood body size and eating behaviour: A structural equation modelling analysis in the Norwegian Mother, Father and Child Cohort Study

作者:

by Tom A. Bond, Tom A. McAdams, Nicole M. Warrington, Laurie J. Hannigan, Espen Moen Eilertsen, Ziada Ayorech, Fartein A. Torvik, George Davey Smith, Deborah A. Lawlor, Eivind Ystrom, Alexandra Havdahl, David M. Evans Background The intergenerational transmission of obesity-related traits could propagate an accelerating cycle of obesity, if parental adiposity causally influences offspring adiposity. The extent to which intergenerational obesity associations are due to such causal effects, as opposed to genetic confounding (inheritance), is unclear. We aimed to establish whether associations between parental peri-pregnancy body mass index (BMI) and offspring birth weight (BW), BMI until 8 years of age, and 8-year-old eating behaviour are due to genetic confounding. Methods and findings Data were from the Norwegian Mother, Father and Child Cohort Study, a prospective population-based birth cohort born between 1999 and 2009 at 50 out of 52 hospital maternity units in Norway. We compared the strength of the associations of maternal pre-pregnancy BMI versus paternal BMI during pregnancy, with offspring outcomes including birth weight and BMI assessed between age 6 months and 8 years of age, and appetite-related eating behaviour traits assessed at age 8 years via the Child Eating Behaviour Questionnaire (CEBQ), adjusting for potential confounders including parity, parental/grandparental language group and parental age, smoking, education and income). We then used an extended children of twins structural equation model (SEM) to quantify the extent to which associations were due to genetic confounding. Up to 85,866 children (51.3% male) were included in linear regression models, whereas SEM models included up to 50,999 children. Maternal BMI was more strongly associated than paternal BMI with offspring BW, but the maternal-paternal difference decreased for offspring BMI after birth. Greater parental BMI was associated with obesity-related offspring eating behaviours. SEM results indicated that genetic confounding did not explain the association between parental BMI and offspring BW, but explained the majority of the association with offspring BMI from 6 months onwards. For 8-year BMI, genetic confounding explained 79% (95% CI [62, 95]; p = 1.9 × 10−12) of the covariance with maternal BMI and 94% (95% CI [72, 113]; p = 2.7 × 10−14) of the covariance with paternal BMI. Limitations of this study include selective recruitment and attrition, potential bias due to parental assortative mating, and that findings may not generalise beyond high-income country settings with high obesity prevalence. Conclusions We found strong evidence that parent–child BMI associations may primarily be due to genetic confounding. When considered alongside prior evidence, this finding may argue against a strong causal effect of maternal or paternal adiposity on childhood adiposity via intrauterine or periconceptional mechanisms.

13.
arXiv (CS.AI) 2026-06-16

ArtNet: A JEPA-Like Articulatory Predictive Framework for Robust Zero-Shot Phoneme Recognition

arXiv:2606.16595v1 Announce Type: cross Abstract: Zero-shot cross-lingual phoneme recognition is often hindered by the fragility of direct acoustic-to-symbol mapping, which is susceptible to language-specific variations. Echoing joint-embedding predictive architecture (JEPA) work in vision, we propose ArtNet, a framework that explores a structured feature prediction task based on articulatory features to enhance acoustic robustness. Specifically, ArtNet integrates an articulatory predictor, designed to extract universal articulatory representations from self-supervised learning (SSL) features, with a variational information bottleneck (VIB) to suppress language-specific variations. Experiments on seven unseen languages demonstrate that ArtNet, particularly when synergized with the proposed vector-space inventory alignment (VSIA) strategy, significantly outperforms competitive baselines, achieving a 20.56\% relative reduction in phoneme error rate (PER) and 7.01\% in phoneme feature error rate (PFER).

14.
arXiv (CS.AI) 2026-06-15

SkillAudit: Ground-Truth-Free Skill Evolution via Paired Trajectory Auditing

arXiv:2606.14239v1 Announce Type: new Abstract: Agent skills are structured procedural packages that guide frozen LLM agents in specialized workflows. Skills rarely remain sufficient after deployment: edge cases, API changes, and deployment constraints become visible only through use, making skill evolution a practical necessity. Existing methods depend on privileged feedback such as held-out validation scores, hidden test outcomes, or environment rewards – signals often unavailable when a practitioner has only a task description and workspace data. We introduce SkillAudit, a framework for evolving agent skills without ground-truth feedback. The key idea is paired trajectory auditing: at each iteration, the same task is executed with and without the candidate skill, isolating how the skill changes agent behavior without external labels. To turn behavioral differences into edit guidance, SkillAudit uses Process-Aligned Contrastive Evaluation (PACE), a cluster of evaluators that maps trajectory divergences to diagnostic signals linked to specific passages in the skill document. A structural verifier, compiled once from the task specification and then fixed, checks task constraints and rolls back harmful updates. SkillAudit routes edits through two pipelines: Refine removes noisy or irrelevant guidance from broadly useful skills, while Repair replaces passages that conflict with the task. Across 89 containerized tasks spanning 8 professional domains, SkillAudit achieves 73.9% average task reward, outperforming an agent without skills (40.9%) and the static expert skill (56.7%). These gains are obtained without accessing hidden tests, reference solutions, or external scoring functions during evolution.

15.
bioRxiv (Bioinfo) 2026-06-19

Sanjeevani: A manually curated anti-cancerous phytochemical database integrated with downstream analysis tools.

Background: Cancer continues to pose a massive global health burden. While plant-derived phytochemicals offer promising therapeutic leads, existing natural product databases often lack cancer specificity, dataset downloadability, and integrated screening tools. Methods: We developed Sanjeevani, an integrative web platform cataloguing 4,823 curated anticancer phytochemicals. Using a balanced dataset of 9,646 molecules, we trained Support Vector Machine (SVM), Random Forest, and K-Nearest Neighbours classifiers using a hybrid feature representation of RDKit descriptors and 2048-bit ECFP4 fingerprints. The platform also integrates AutoDock Vina for web-based molecular docking for binding affinity, poses prediction and ADMET-AI for pharmacokinetics estimation. Results: The SVM model demonstrated the strongest predictive capability, achieving a top test accuracy of 0.966 and a ROC-AUC of 0.992. Benchmarking across five docking tools confirmed that AutoDock Vina successfully balanced computational automation with literature-consistent binding affinity replication. The final architecture provides rapid interactive 2D/3D visualizations integrated with downstream analysis tools. Conclusion: Sanjeevani provides an open-access, one-stop pipeline that bridges the gap between raw natural product data and actionable computational screening, accelerating natural product-based oncology drug discovery.

16.
arXiv (CS.LG) 2026-06-11

Analytic Bijections for Smooth and Interpretable Normalizing Flows

arXiv:2601.10774v2 Announce Type: replace Abstract: A key challenge in normalizing flows is finding expressive invertible scalar bijections. Existing approaches face trade-offs: affine transformations are smooth and analytically invertible but lack expressivity; monotonic splines offer local control but are only piecewise smooth and act on bounded domains; residual flows achieve smoothness but need numerical inversion. We introduce three families of analytic bijections that are globally smooth ($C^\infty$), defined on all of $\mathbb{R}$, and analytically invertible in closed form, combining the favorable properties of prior approaches. Beyond serving as drop-in replacements in coupling flows, where they match or exceed spline performance, we develop radial flows: a novel architecture using direct parametrization that transforms the radial coordinate while preserving angular direction. Radial flows exhibit exceptional training stability, produce geometrically interpretable transformations, and on targets with radial structure can achieve comparable quality to coupling flows with $1000\times$ fewer parameters. We provide comprehensive evaluation on 1D and 2D benchmarks, and demonstrate applicability to higher-dimensional physics problems through experiments on $\phi^4$ lattice field theory, where our bijections outperform affine baselines and enable problem-specific designs that address mode collapse.

17.
arXiv (CS.LG) 2026-06-17

On Randomized Algorithms in Online Strategic Classification

arXiv:2602.06257v2 Announce Type: replace Abstract: Online strategic classification studies settings in which agents strategically modify their features to obtain favorable predictions. For example, given a classifier that determines loan approval based on credit scores, applicants may open or close credit cards and bank accounts to obtain a positive prediction. The learning goal is to achieve low mistake or regret bounds despite such behavior. While randomized algorithms have the potential to offer advantages to the learner in strategic settings, they have been largely underexplored. In the realizable setting, no lower bound is known for randomized algorithms, and existing lower bound constructions for deterministic learners can be circumvented by randomization. In the agnostic setting, the best known regret upper bound is $O(T^{3/4}\log^{1/4}T|\mathcal H|)$, which is far from the standard online learning rate of $O(\sqrt{T\log|\mathcal H|})$. In this work, we provide refined bounds for online strategic classification in both settings; our bounds depend on the Littlestone dimension $\mathrm{Ldim}(\mathcal H)$ of the hypothesis class $\mathcal H$ and the maximum degree $\Delta$ of the manipulation graph. In the realizable setting, we extend, for $T > \mathrm{Ldim}(\mathcal H) \Delta^2$, the existing lower bound $\Omega(\mathrm{Ldim}(\mathcal H) \Delta)$ for deterministic learners to all learners. This yields the first lower bound that applies to randomized learners. We then provide the first randomized learner that improves the known (deterministic) upper bound of $O(\mathrm{Ldim}(\mathcal H) \cdot \Delta \log \Delta)$. In the agnostic setting, we give an improper randomized learner that improves the regret upper bound to $O(\sqrt{T\log|\mathcal H|})$, matching the standard online learning rate. We also show a larger lower bound for all proper learning rules, demonstrating that improperness is necessary to achieve the optimal rate.

18.
arXiv (CS.AI) 2026-06-16

Learn from Your Mistakes: Tree-like Self-Play for Secure Code LLMs

arXiv:2606.03489v2 Announce Type: replace-cross Abstract: While Large Language Models (LLMs) excel in code generation, they remain prone to replicating subtle yet critical vulnerabilities endemic to their training data. Current alignment techniques, such as Supervised Fine-Tuning (SFT) and Reinforcement Learning (RL), typically apply coarse-grained optimization at the sequence level. This approach often fails to address the localized nature of security flaws, where a single incorrect token choice can compromise an entire program. To bridge this gap, we introduce Tree-like Self-Play (TSP), a framework that reframes secure code generation as a fine-grained sequential decision process. Unlike standard methods that blindly maximize likelihood, TSP constructs a decision tree where the model explores branching trajectories–generating both secure "golden paths" and vulnerable variants. By treating code generation as a self-play game, the model learns to strictly discriminate against its own localized errors. This provides a dense, on-policy learning signal that forces self-correction precisely at the critical decision nodes where vulnerabilities typically emerge. Our experiments demonstrate that TSP fundamentally enhances model reliability. In Python security benchmarks, TSP boosts CodeLlama-7B's pass rate (SPR@1) to 75.8%, significantly outperforming SFT (57.0%) and unstructured self-play baselines. Crucially, TSP induces robust out-of-distribution generalization: the model not only reduces vulnerabilities in unseen categories (CWEs) by 24.5% but also successfully transfers security principles learned from C/C++ to diverse languages, including Python, Go, and JavaScript. This suggests that TSP does not merely memorize patches, but internalizes abstract, language-agnostic security logic.

19.
arXiv (CS.LG) 2026-06-19

Learning to Emulate Chaos: Adversarial Optimal Transport Regularization

arXiv:2604.21097v2 Announce Type: replace-cross Abstract: Chaos arises in many complex dynamical systems, from weather to power grids, but is difficult to accurately model with data-driven methods such as machine learning emulators. While emulators are promising tools for accelerating simulations and solving inverse problems, they still struggle to learn chaotic dynamics, where sensitivity to initial conditions renders exact long-term forecasts infeasible, especially given noisy data. Recent work instead trains emulators to match the statistical properties of chaotic attractors, but these approaches often rely on handcrafted summary statistics or large, diverse multi-environment datasets. In this work, we propose a family of adversarial optimal transport objectives that can jointly learn high-quality summary statistics and a physically consistent emulator from a single noisy trajectory. We theoretically analyze and experimentally validate a Sinkhorn divergence formulation (2-Wasserstein) and a WGAN-style dual formulation (1-Wasserstein) of our approach. Numerical experiments across a variety of chaotic systems, including ones with high-dimensional spatiotemporal chaos, show that emulators trained using our proposed objectives have significantly improved long-term statistical fidelity.

20.
medRxiv (Medicine) 2026-06-17

The interaction between chronic hepatitis B (CHB) and Metabolic dysfunction-associated steatotic liver disease (MASLD) in a diverse central London population

Introduction: The overlap between chronic hepatitis B (CHB) and metabolic dysfunction-associated steatotic liver disease (MASLD) is an emerging global health challenge. We investigated the impact of MASLD and metabolic comorbidity in a diverse London viral hepatitis clinic. Methods: This retrospective cross-sectional study (May 2018-Feb 2024) included adults with CHB having controlled attenuation parameter (CAP) measurements. MASLD was defined as CAP >264 dB/m plus [≥]1 cardiometabolic factor (CMF). We used univariable and multivariable models to examine MASLD's relationship with liver stiffness and hepatitis B viral load (HBV VL). Results: Among 323 individuals (67% male, median age 36), most were from Black (35%) or non-white British/Irish (29%) backgrounds. Overall, 64% had [≥]1 CMF, and 20% had MASLD. The CHB/MASLD group was significantly older (median 43 vs 35 years, p

21.
arXiv (CS.AI) 2026-06-15

Closing the Reflection Gap: A Free Calibration Bonus for Agentic RL

作者:

arXiv:2606.14211v1 Announce Type: new Abstract: LLMs are increasingly deployed as agents that interact with external environments and observe feedback such as execution results, error messages, and tool outputs. A well-functioning agent should be able to leverage this feedback to accurately assess its own performance. Yet we find a persistent reflection gap: LLM agents tend to mis-assess their own outputs after observing concrete environment feedback – even for questions they correctly answered – and standard RL barely helps due to a credit-assignment mismatch. To close this gap, we propose RefGRPO, a simple yet effective fix that augments standard RL algorithms with two key ingredients: a free calibration bonus computed by contrasting the agent's own reflection with the actual outcome (requiring no additional reward model, LLM judge, or external annotation), and a dynamic schedule on its coefficient. Compared to standard RL baselines, our method simultaneously improves reflection calibration (e.g., reduces underconfidence rate $44.4\% \to 7.7\%$) and task accuracy (e.g., $75.1\% \to 76.5\%$) on text-to-SQL across five benchmarks. The resulting calibrated reflection turns the agent into its own verifier grounded in environment feedback, which further enables (i) better self-improvement that uses reflections as pseudo-rewards without outcome supervision, and (ii) more effective test-time selective prediction by committing only to rollouts flagged as correct.

22.
arXiv (quant-ph) 2026-06-12

Multi-entropy in heavy local quenches

arXiv:2606.12526v1 Announce Type: cross Abstract: We study the time evolution of tripartite entanglement in heavy local quenches in two-dimensional holographic conformal field theories. Our diagnostic is the genuine multi-entropy of adjacent intervals, computed from both bulk and boundary perspectives. A perturbative bulk analysis shows that the first-order small-mass perturbation around the vacuum geodesic network cancels identically at any time after the quench. In the fully back-reacted geometry, a vacuum-subtracted genuine multi-entropy arises from a mismatch between the winding selected by the trivalent geodesic network and the windings selected independently by the pairwise geodesics. In the sharp quench limit, the time dependence of genuine multi-entropy is kinematically fixed to logarithms of rational functions of time and is independent of the heavy operator dimension. The CFT calculation reproduces the same formula within the heavy-light vacuum block approximation, where the branch choice in the heavy-background uniformization map corresponds to the winding selection in the bulk. These results indicate that, in this setup, the genuine multi-entropy is controlled by global saddle selection, rather than by a local energy response or quasiparticle propagation.

23.
arXiv (CS.LG) 2026-06-11

Re-evaluating Confidence Remasking in Masked Diffusion Language Models

arXiv:2606.12232v1 Announce Type: new Abstract: Masked diffusion language models (dLLMs) have recently emerged as a competitive alternative to autoregressive language models, with the promise of faster inference via parallel token generation. A notable limitation of the masked formulation, however, is that once a token has been unmasked it can no longer be revised, leaving dLLMs vulnerable to early sampling mistakes. To address this, a growing body of work has sought to extend masked dLLMs with self-correcting (remasking) capabilities. One appealing subset of these methods does so in a training-free, post-hoc manner based on token confidences, with encouraging early reported results. In this work, we revisit the empirical evaluation of a representative post-hoc remasking method, WINO [Hong et al., 2026], and find that under standard decoding settings (shorter block lengths) it brings little-to-no benefit over confidence-based unmasking alone [Wu et al., 2025]. Extending the evaluation to non-greedy decoding, we find that while confidence-based remasking can mitigate errors introduced by increased stochasticity to some extent, it also exacerbates the diversity collapse previously reported for confidence-based unmasking. Overall, our results show that the benefits of post-hoc confidence-based remasking are highly setting-dependent, underscoring the need for a more comprehensive evaluation framework.

24.
arXiv (CS.AI) 2026-06-16

Proximal Policy Optimization for Amortized Discrete Sampling

arXiv:2606.15793v1 Announce Type: cross Abstract: This paper explores policy gradient algorithms for training stochastic policies to sample from structured discrete probability distributions under the Generative Flow Network (GFlowNet) framework. Building on extensive theoretical connections between GFlowNets and entropy-regularized reinforcement learning, we derive equivalents of standard policy gradient algorithms for training GFlowNets, as well as experimentally explore their various methodological aspects, including baseline training and advantage estimation. Most importantly, our work is the first to derive and successfully apply proximal policy optimization to GFlowNets, showing its improved convergence speed and data efficiency compared to standard GFlowNet training objectives on benchmarks ranging from synthetic energies to molecular graph generation.

25.
arXiv (CS.AI) 2026-06-12

Mining Architectural Quality Under Agentic AI Adoption: A Causal Study of Java Repositories

arXiv:2606.13298v1 Announce Type: cross Abstract: AI coding tools are now used by a majority of developers, and agentic use of these tools has popularized the practice colloquially called "vibe coding". Yet causal evidence on their effect on software architecture is scarce. Prior causal work has measured code-level outcomes (complexity, static analysis warnings); whether such degradation propagates to architecture-level outcomes remains unknown. We mine 151 open-source Java repositories, 74 with detectable agentic AI adoption (identified via configuration files and Co-Authored-By commit trailers) and 77 propensity-matched controls, across a 13-month per-repository window yielding 1,811 monthly Arcan snapshots. We estimate the causal effect of adoption on architectural smell density (ASD) with a staggered difference-in-differences design and the Borusyak imputation estimator, applying a causal design recently used for code-level metrics to the architecture level. Total smell counts are essentially unchanged (+1.1%, p = 0.82) while lines of code grow +12.8% (p = 0.003); the resulting 6.7% ASD decline (p = 0.004) is therefore a denominator effect rather than an architectural improvement. Per-type estimates and robustness checks (wild cluster bootstrap, Lee bounds, stale-observation sensitivity) corroborate the pattern; pre-trends are flat (Wald p = 0.90), consistent with parallel trends. Density-normalized outcomes can mislead when treatment affects system size: raw counts and explicit decomposition are required for causal mining studies of AI tool adoption. The complete replication package, including the curated 151-repository monthly panel, is publicly available.