As large language models (LLMs) continue to grow in size, distributed inference has become increasingly important. Model-parallel strategies must now efficiently scale not only across multiple GPUs but also across multiple nodes. In this work, we present a detailed performance study of multi-node distributed inference using LLMs on GPU-based supercomputers. We conduct experiments with several state-of-the-art inference engines alongside YALIS, a research-oriented prototype engine designed for controlled experimentation. We analyze the strong-scaling behavior of different model-parallel schemes and identify key bottlenecks. Because all-reduce operations are a common performance bottleneck, we develop NVRAR, a hierarchical all-reduce algorithm based on recursive doubling with NVSHMEM. NVRAR achieves up to 1.9$\times$-3.6$\times$ lower latency than NCCL for message sizes between 128 KB and 2 MB on HPE Slingshot and InfiniBand interconnects. Integrated into YALIS, NVRAR achieves up to a 1.72$\times$ reduction in end-to-end batch latency for the Llama 3.1 405B model in multi-node decode-heavy workloads using tensor parallelism.

Large language models show promise for financial decision-making, yet deploying them as autonomous trading agents raises fundamental challenges: how to adapt instructions when rewards arrive late and obscured by market noise, how to synthesize heterogeneous information streams into coherent decisions, and how to bridge the gap between model outputs and executable market actions. We present ATLAS (Adaptive Trading with LLM AgentS), a unified multi-agent framework that integrates structured information from markets, news, and corporate fundamentals to support robust trading decisions. Within ATLAS, the central trading agent operates in an order-aware action space, ensuring that outputs correspond to executable market orders rather than abstract signals. The agent can incorporate feedback while trading using Adaptive-OPRO, a novel prompt-optimization technique that dynamically adapts the prompt by incorporating real-time, stochastic feedback, leading to increasing performance over time. Across regime-specific equity studies and multiple LLM families, Adaptive-OPRO consistently outperforms fixed prompts, while reflection-based feedback fails to provide systematic gains.

Recent advances in large language models (LLMs) have significantly improved automated code generation. While existing approaches have achieved strong performance at the function and file levels, real-world software engineering requires reasoning over entire repositories, including cross-file dependencies, evolving execution environments, and global semantic consistency. This challenge has led to the emergence of Repository-Level Code Generation (RLCG), where models must retrieve, organize, and utilize repository-scale context to generate coherent and executable code changes. To address these challenges, Retrieval-Augmented Generation (RAG) has become an increasingly important paradigm for repository-level code intelligence. In this survey, we present a comprehensive review of Retrieval-Augmented Code Generation (RACG), with a particular focus on repository-level approaches. Rather than viewing RACG as a static ``retrieve-then-generate'' pipeline, we characterize it as a coupled and evolving process involving context construction, retrieval optimization, generation, and environment interaction. We organize existing methods through a unified analytical framework spanning retrieval substrate, control regime, and evaluation setting. Based on this framework, we systematically examine retrieval strategies, graph-based and non-graph-based retrieval paradigms, training-driven optimizations, and autonomous agent architectures. We further summarize widely used datasets, benchmarks, and system configurations, and discuss key challenges including scalability, reliability, efficiency, and the necessity boundary between RACG and long-context LLMs. Through this survey, we aim to provide a structured understanding of the rapidly evolving RACG landscape and highlight promising directions for future AI-powered software engineering research.

Large Language Model (LLM)-based coding agents have shown promising results on coding benchmarks, but their effectiveness on systems code remains underexplored. Due to the size and complexities of systems code, making changes to a systems codebase requires researching about many pieces of context, derived from the large codebase and its massive commit history, before making changes. Inspired by the recent progress on deep research agents, we design the first deep research agent for code, called Code Researcher, and apply it to the problem of generating patches to mitigate crashes reported in systems code. Code Researcher performs multi-step reasoning about semantics, patterns, and commit history of code to retrieve all relevant context from the codebase and its commit history. We evaluate Code Researcher on kBenchSyz, a benchmark of Linux kernel crashes, and show that it significantly outperforms strong baselines, achieving a crash-resolution rate (CRR) of 48%, compared to 31.5% by SWE-agent and 31% by Agentless, using OpenAI's GPT-4o model. Scaling up sampling budget to 10 trajectories increases Code Researcher's CRR to 54%. Code Researcher is also robust to model choices, reaching 67% with the newer Gemini 2.5-Flash model. Through another experiment on an open-source multimedia software, we show the generalizability of Code Researcher and also conduct ablations. Our experiments highlight the importance of global context gathering and multi-faceted reasoning for large codebases.

Speaker anonymization systems hide the identity of speakers while preserving other information such as linguistic content and emotions. To evaluate their privacy benefits, attacks in the form of automatic speaker verification (ASV) systems are employed. In this study, we assess the impact of intra-speaker linguistic content similarity in the attacker training and evaluation datasets, by adapting BERT, a language model, as an ASV system. On the VoicePrivacy Attacker Challenge datasets, our method achieves a mean equal error rate (EER) of 35%, with certain speakers attaining EERs as low as 2%, based solely on the textual content of their utterances. Our explainability study reveals that the system decisions are linked to semantically similar keywords within utterances, stemming from how LibriSpeech is curated. Our study suggests reworking the VoicePrivacy datasets to ensure a fair and unbiased evaluation and challenge the reliance on global EER for privacy evaluations.

The advancement of large language models (LLMs) has created a competitive landscape for AI-assisted programming tools. This study evaluates two leading models: ChatGPT 03-mini and DeepSeek-R1 on their ability to solve competitive programming tasks from Codeforces. Using 29 programming tasks of three levels of easy, medium, and hard difficulty, we assessed the outcome of both models by their accepted solutions, memory efficiency, and runtime performance. Our results indicate that while both models perform similarly on easy tasks, ChatGPT outperforms DeepSeek-R1 on medium-difficulty tasks, achieving a 54.5% success rate compared to DeepSeek 18.1%. Both models struggled with hard tasks, thus highlighting some ongoing challenges LLMs face in handling highly complex programming problems. These findings highlight key differences in both model capabilities and their computational power, offering valuable insights for developers and researchers working to advance AI-driven programming tools.

The multi-armed bandits (MAB) framework is a widely used approach for sequential decision-making, where a decision-maker selects an arm in each round with the goal of maximizing long-term rewards. In many practical applications, such as personalized medicine and recommendation systems, contextual information is available at the time of decision-making, rewards from different arms are related rather than independent, and feedback is provided in batches. We propose a novel semi-parametric framework for batched bandits with covariates that incorporates a shared parameter across arms. We leverage the single-index regression (SIR) model to capture relationships between arm rewards while balancing interpretability and flexibility. Our algorithm, Batched single-Index Dynamic binning and Successive arm elimination (BIDS), employs a batched successive arm elimination strategy with a dynamic binning mechanism guided by the single-index direction. We consider two settings: one where a pilot direction is available and another where the direction is estimated from data, deriving theoretical regret bounds for both cases. When a pilot direction is available with sufficient accuracy and the number of arms $K$ is fixed, our approach achieves minimax-optimal rates (with $d = 1$) for nonparametric batched bandits, circumventing the curse of dimensionality. Extensive experiments on simulated and real-world datasets demonstrate the effectiveness of our algorithm compared to the nonparametric batched bandit method introduced by \cite{jiang2025batched}.

The human brain receives nutrients and oxygen through an intricate network of blood vessels. Pathology affecting small vessels, at the mesoscopic scale, represents a critical vulnerability within the cerebral blood supply and can lead to severe conditions, such as Cerebral Small Vessel Diseases. The advent of 7 Tesla MRI systems has enabled the acquisition of higher spatial resolution images, making it possible to visualise such vessels in the brain. However, the lack of publicly available annotated datasets has impeded the development of robust, machine learning-driven segmentation algorithms. To address this, the SMILE-UHURA challenge was organised. This challenge, held in conjunction with the ISBI 2023, in Cartagena de Indias, Colombia, aimed to provide a platform for researchers working on related topics. The SMILE-UHURA challenge addresses the gap in publicly available annotated datasets by providing an annotated dataset of Time-of-Flight angiography acquired with 7T MRI. This dataset was created through a combination of automated pre-segmentation and extensive manual refinement. In this manuscript, sixteen submitted methods and two baseline methods are compared both quantitatively and qualitatively on two different datasets: held-out test MRAs from the same dataset as the training data (with labels kept secret) and a separate 7T ToF MRA dataset where both input volumes and labels are kept secret. The results demonstrate that most of the submitted deep learning methods, trained on the provided training dataset, achieved reliable segmentation performance. Dice scores reached up to 0.838 $\pm$ 0.066 and 0.716 $\pm$ 0.125 on the respective datasets, with an average performance of up to 0.804 $\pm$ 0.15.

In the field of parameterized complexity theory, the study of graph width measures has been intimately connected with the development of width-based model checking algorithms for combinatorial properties on graphs. In this work, we introduce a general framework to convert a large class of width-based model-checking algorithms into algorithms that can be used to test the validity of graph-theoretic conjectures on classes of graphs of bounded width. Our framework is modular and can be applied with respect to several well-studied width measures for graphs, including treewidth and cliquewidth. As a quantitative application of our framework, we prove analytically that for several long-standing graph-theoretic conjectures, there exists an algorithm that takes a number $k$ as input and correctly determines in time double-exponential in $k^{O(1)}$ whether the conjecture is valid on all graphs of treewidth at most $k$. These upper bounds, which may be regarded as upper-bounds on the size of proofs/disproofs for these conjectures on the class of graphs of treewidth at most $k$, improve significantly on theoretical upper bounds obtained using previously available techniques.

This paper proposes an automatic image correction method for portrait photographs, which promotes consistency of facial skin color by suppressing skin color changes due to background colors. In portrait photographs, skin color is often distorted due to the lighting environment (e.g., light reflected from a colored background wall and over-exposure by a camera strobe), and if the photo is artificially combined with another background color, this color change is emphasized, resulting in an unnatural synthesized result. In our framework, after roughly extracting the face region and rectifying the skin color distribution in a color space, we perform color and brightness correction around the face in the original image to achieve a proper color balance of the facial image, which is not affected by luminance and background colors. Unlike conventional algorithms for color correction, our final result is attained by a color correction process with a guide image. In particular, our guided image filtering for the color correction does not require a perfectly-aligned guide image required in the original guide image filtering method proposed by He et al. Experimental results show that our method generates more natural results than conventional methods on not only headshot photographs but also natural scene photographs. We also show automatic yearbook style photo generation as an another application.

Compositional score-based approaches to simulation-based inference (SBI) approximate the posterior over a shared parameter given $n$ independent observations by aggregating individually learned posterior scores: currently, there are two main propositions of such methods (Geffner et al. (2023), Linhart et al. (2026)). As the resulting composite score does not correspond to the score of any distribution along the forward diffusion path of the true multi-observation posterior, sampling from it via a reverse SDE leads to an irreducible bias. Annealed Langevin dynamics provides a principled alternative: it treats the composite score as the genuine score of a sequence of tractable bridging densities and samples from them in succession. When properly tuned, it could lead to a controllable bias. However, its hyperparameters, namely step sizes, the number of steps per level, and the number of annealing levels, have so far been chosen empirically. We derive Wasserstein bounds for annealed Langevin with approximate scores and translate them into explicit decision rules for these hyperparameters that guarantee a prescribed sampling accuracy, while highlighting different theoretical aspects of each composite score formulation. In the Gaussian setting, we obtain closed-form expressions for all relevant quantities and prove that the bridging densities of Linhart et al. (2026) consistently admit larger step sizes and require fewer total Langevin steps than those of Geffner et al. (2023). Furthermore, we show empirically that the tuning obtained in the Gaussian setting generalizes to more complex problems, thus providing a well-understood and theoretically grounded starting point for practitioners using compositional score-based approaches.

A retinal vessel analysis is a procedure that can be used as an assessment of risks to the eye. This work proposes an unsupervised multimodal approach that improves the response of the Frangi filter, enabling automatic vessel segmentation. We propose a filter that computes pixel-level vessel continuity while introducing a local tolerance heuristic to fill in vessel discontinuities produced by the Frangi response. This proposal, called the local-sensitive connectivity filter (LS-CF), is compared against a naive connectivity filter to the baseline thresholded Frangi filter response and to the naive connectivity filter response in combination with the morphological closing and to the current approaches in the literature. The proposal was able to achieve competitive results in a variety of multimodal datasets. It was robust enough to outperform all the state-of-the-art approaches in the literature for the OSIRIX angiographic dataset in terms of accuracy and 4 out of 5 works in the case of the IOSTAR dataset while also outperforming several works in the case of the DRIVE and STARE datasets and 6 out of 10 in the CHASE-DB dataset. For the CHASE-DB, it also outperformed all the state-of-the-art unsupervised methods.

As a rapidly emerging interdisciplinary field that intrinsically integrates microwave and photonics, microwave photonics (MWP) provides disruptive solutions to overcome the fundamental bandwidth of conventional electronic systems. By exploiting the inherently ultra-wide bandwidth and low-loss characteristics of photonic technologies, MWP enables the generation, transmission, processing, and detection of microwave, millimeter-wave, and terahertz signals. Representative breakthroughs include fully photonic microwave radar systems, photonic analog-to-digital converters with bandwidth up to 320 GHz, and photonic wireless communication systems achieving data rate as high as 616 Gbit/s. Meanwhile, the rapid growth of artificial intelligence (AI) is reshaping scientific research, engineering, and daily life in unprecedented ways, such as AI for science/engineering and AI co-scientist/assistant. Correspondingly, AI is profoundly reshaping MWP in all aspects, ranging from signal generation, transmission to signal processing and detection. AI has revolutionized the design, simulation, fabrication, testing, deployment, and maintenance of MWP systems, delivering autonomous operation and exceptional efficiency beyond traditional systems. Motivated by these developments, this Review Paper provides the first comprehensive overview of AI-enabled MWP, systematically summarizing the state-of-the-art advances and presenting insights for both the academic community and the broader public.

Low-rank adaptation (LoRA) has emerged as a powerful tool for parameter-efficient fine-tuning of large language models (LLMs). This paper studies LoRA under a federated learning setting, enabling collaborative fine-tuning across clients while preserving parameter efficiency. We focus on a highly heterogeneous regime in which clients share only partial structure and a substantial subset may be contaminated. We propose Collaborative Low-rank Alignment and Identifiable Recovery (CLAIR), a contamination-aware framework that relies only on preliminary local estimators. Its formulation applies broadly, from linear regression to neural network and LLM modules, whenever local adaptation can be represented by matrix-valued updates. CLAIR recovers the shared LoRA subspace and detects contaminated clients via a structured low-rank plus block-sparse decomposition. We prove exact recovery of the shared LoRA subspace in the noiseless case, stable recovery under preliminary estimation error, and consistent collaborative-set recovery under mild separation conditions. We further quantify the gain from CLAIR refinement: it reduces off-subspace estimation error through cross-client averaging while preserving client-specific variation within the shared LoRA subspace, thus improves over local fine-tuning whenever this oracle gain outweighs the costs of subspace estimation and benign-client heterogeneity. Empirically, we demonstrate the benefits of CLAIR by fine-tuning a Transformer architecture on a text-copying task. The results show accurate contamination detection and improved benign-client performance compared with local fine-tuning and non-robust federated averaging.

In this work, we present quantum reinforcement learning (RL) as a solution strategy for process synthesis problems. Building on our prior work, we develop a generalized framework that formally poses process synthesis as a Markov decision process and introduces quantum-enhanced RL algorithms to solve it with improved scalability. Earlier implementations of quantum-based RL for process synthesis were limited by qubit requirements, which scaled poorly with problem complexity. This work overcomes this challenge by introducing state encoding algorithms to decouple qubit requirements from problem size. A classical RL-based solution strategy is used as a baseline to benchmark the quantum algorithms under identical training conditions. All algorithms are evaluated across a flowsheet synthesis problem of increasing unit counts to analyze their performance and scalability. Results show that all approaches are capable of identifying the optimal flowsheet designs in small design spaces. For moderate-scale unit counts, quantum approaches demonstrate competitive performance on a per-episode basis and improved efficiency on a per-parameter basis versus the classical RL benchmark. This work provides a foundation for future quantum computing applications within process systems engineering, establishes a controlled benchmark for comparing classical and quantum algorithms, and shows that the proposed quantum variants remain competitive for the process synthesis problem examined in this work.

In this work we present an efficient and practically implementable approach for the application of reinforcement learning (RL)-based control in chemical process systems. This is an area that has yet to widely adopt RL-based control largely due to inherent challenges in trusting RL algorithms and the time-consuming process of training reliable agents. To address these challenges, we leverage a class of RL algorithms termed Y-wise Affine Neural Network (YANN)- RL, which we have developed in our prior work (Braniff and Tian, 2025a). By strategically initializing actor and critic networks YANN-RL algorithms provide confident and interpretable starting points within control schemes. We apply this RL-based control approach to three different process engineering case studies publicly available on the PC-Gym library (Bloor et al., 2026): (i) a continuous stirred tank reactor (CSTR), (ii) a four-tank system, and (iii) a multistage extraction column. Our approach is compared to several popular RL algorithms (PPO, SAC, DDPG, and TD3) and is benchmarked against nonlinear model predictive control (NMPC). These case studies demonstrate that YANN-RL can greatly reduce the training time and data needed, can be deployed with confidence for chemical process systems, and can approach the performance of NMPC without the knowledge of a full nonlinear model.

Public events on social media generate large volumes of discussion whose collective dynamics carry direct value for opinion forecasting and crisis response. Capturing how these dynamics evolve across an event's lifecycle requires organizing fragmented posts into event-level time series. Existing datasets cover only a small number of events within a single category, and typically discard the interaction structure between posts when constructing time series, which restricts both transfer across event types and controlled study of how interactions shape the resulting collective dynamics. We present SURGE, a multi-event social media benchmark that pairs event-level time series with aligned text and interaction structure linking posts within an event. SURGE is built through an automated pipeline that produces calendar-aligned time series at three temporal granularities, covering 67 events and more than 800K posts across five event categories. Each time bin is paired with flat and structured textual views derived from the same selected posts, enabling controlled evaluation of whether social interaction structure affects forecasting behavior. On top of SURGE we define benchmark protocols for numerical-only forecasting, text-augmented forecasting, high-interaction evaluation, and leave-one-category-out generalization. Experiments with representative time-series and multimodal forecasting models reveal three properties of the benchmark: a strong local-persistence regime in which naive baselines remain hard to beat under absolute error, limited transfer of existing text-augmented forecasters to event-driven social-media data, and increased difficulty on reply-dense periods that aggregate metrics tend to obscure. We further include a lightweight structure-aware probe as a reference implementation, illustrating how SURGE can support interaction-aware forecasting research.

An accurate assessment of a model's complexity is crucial for topics such as interpretation, generalization, and model selection. However, most existing complexity measures either rely on heuristic assumptions or are computationally prohibitive. In this paper, we present a mathematically rigorous yet easy-to-compute measure of model complexity that is based on the similarities between the model gradients across inputs. It is thus well-defined for any parametric model, but also for kernel-based non-parametric models. We prove that our measure of complexity generalizes model-specific complexity measures such as polynomial degree (for polynomial regression), kernel length scale (for Matérn kernels), number of neighbors (for k-nearest neighbors), number of splits (for decision trees), and number of trees (for random forests). We also use our measure to obtain new insights into the double descent phenomenon for random Fourier features, random forests, neural networks, and gradient boosting.

Modern DNNs are repeatedly fine-tuned to incorporate new data and functionality. This evolutionary workflow introduces a security risk when updated data cannot be fully trusted, as adversaries may implant Trojans during fine-tuning. We present MIST, a Trojan detection approach that analyzes how a model's internal representations change during fine-tuning. Rather than attempting to reconstruct trigger conditions, MIST characterizes benign model evolution using pre-activation spectra and flags updates whose spectral deviations are inconsistent with this reference. This framing treats Trojan detection as a regression problem over model updates. An empirical evaluation across four datasets and eight Trojan attacks shows that spectral distances reliably distinguish Trojaned updates from clean fine-tuning. MIST outperforms state-of-the-art detection accuracy after a single update, without requiring any knowledge about the poisoned data or the trigger, and remains effective under multi-step benign evolution, with graceful and bounded degradation. These results indicate that spectral evolution provides a stable and assumption-light signal for detecting malicious model updates.

This paper establishes and proves complexity results for entailment for cumulative propositional dependence logic and for cumulative propositional logic with team semantics. As recently shown, cumulative logics are famously characterised by System~C and exactly captured by the cumulative models of Kraus, Lehmann and Magidor. This gives rise to the entailment problem via relational models, which is specifically considered here.

Communication enables coordination in multi-agent reinforcement learning (MARL), but many real-world applications, e.g., search-and-rescue with drone swarms, operate under severe bandwidth constraints. Many communication architectures still expose a coupled bottleneck in which a shared latent representation is used for both policy execution and inter-agent communication. Consequently, reducing message size directly limits the policy's latent space, often leading to significant performance degradation. We address this with two contributions. First, we introduce $β$, a normalised per-agent bandwidth budget that unifies sparsity, rounds, and message dimension into a single comparable constraint. Second, we provide SLIM, a minimal architecture that decouples the communication pathway from the policy's latent representation, allowing us to isolate the effect of bandwidth from the effect of policy capacity while benefiting from in-step communication. We evaluate our method on several partially-observable MARL benchmarks, where communication is essential. Our approach achieves state-of-the-art performance and exhibits scalability and robustness under limited communication, with only marginal degradation as bandwidth is reduced.

Materials discovery and biomedical translation increasingly require models that can reason across composition, processing, structure, biological response, manufacturability, safety, and governance constraints. Existing materials and biomedical data ecosystems are powerful but remain poorly coupled for AI-guided discovery. Here we present AIMBio, a conceptual framework for an AI-native, FAIR, and governance-aware decision layer that links materials provenance, biomedical context, knowledge graphs, uncertainty-aware machine learning, and human-in-the-loop active learning. The framework formulates biomedical-materials discovery as constrained multi-objective optimization under uncertainty and introduces practical requirements for metadata, model documentation, risk-tiered governance, evaluation metrics, and phased implementation. To make the roadmap testable, we add a minimum viable prototype specification and a worked pilot for AI-guided nanomaterials for drug delivery. AIMBio is positioned as exploratory and preclinical discovery infrastructure, not as clinical decision-support software; any clinical or regulated-device use would require separate validation, change control, and regulatory review. The central contribution is a publishable platform blueprint for converting fragmented materials and biomedical records into auditable, experimentally actionable, and translationally responsible discovery workflows.

A recent trend is to leverage machine learning models to improve the evolutionary design and optimization process. We propose a novel transformer-based mutation operator for Cartesian genetic programming (CGP) for the automated design of approximate arithmetic circuits. We introduce a hybrid scheme for CGP in which the proposed mutation operator is switched with the standard mutation operator to prevent stagnation of the circuit approximation process. We also develop a new training scheme for the underlying transformer that utilizes training vectors composed of thousands of CGP chromosomes representing various approximate multipliers. For several target error constraints, the approximate multipliers evolved with CGP utilizing the transformer-based mutation achieve better trade-offs than the highly optimized designs available in the state-of-the-art EvoApproxLib library of approximate circuits. Although both training and evolutionary processes are computationally demanding, they appear to be necessary steps for improving existing approximate circuits and producing new, potentially patentable circuit designs.

Gaussian processes (GPs) offer a principled probabilistic model over functions, but exact inference is restricted to the linear-Gaussian regime. We establish an explicit equivalence between GPs and a class of linear diffusion models, recasting predictive sampling as an ODE with closed-form Gaussian dynamics and a likelihood-dependent guidance term that admits a simple Monte Carlo approximation. In the linear-Gaussian setting, we recover standard GP conditioning exactly; beyond conjugacy, the same machinery handles any conditioning statement admitting point-wise likelihood evaluation -- including non-linear physics, and, for the first time, natural language via large language models. Whitening isolates the irreducible non-Gaussian dynamics, minimising Wasserstein-2 transport cost and eliminating numerical stiffness. The result is a general-purpose GP inference scheme requiring no bespoke derivations. Together, these results provide a general mechanism for incorporating the full richness of real-world knowledge as conditioning information, opening a new frontier for the probabilistic modelling of real-world problems.

Generative artificial intelligence now synthesizes photorealistic imagery, audio, and video at a cost that defeats traditional forensic intuition. The legal consequences span three regimes studied so far in isolation: international operational law, domestic procedure, and product regulation. This article presents a unified evidentiary framework that maps cryptographic content provenance, robust statistical watermarking, and zero knowledge attestation to the proof requirements of each regime. We define a five tier threat model spanning naive regeneration, adversarial laundering, cross model regeneration, active watermark removal, and insider provenance forgery. We release a public benchmark of 12000 generated items across image, audio, and video modalities under six laundering pipelines for 72000 evaluation samples. We evaluate four representative schemes and report true positive rate at fixed false positive rate, robustness area under the curve, computational overhead, and a regime conditioned legal sufficiency score. We translate empirical detection bounds into legal sufficiency thresholds for command decisions under the law of armed conflict, for criminal and civil admissibility under domestic procedure, and for persistence audits under the European Union Artificial Intelligence Act and analogous regimes. The result is a reproducible reference pipeline, a public benchmark, and model annexes that lawyers, engineers, and operators can deploy together.

We establish maximal concentration bounds for the iterates generated by stochastic approximation algorithms with general step sizes, where the noise has a finite-state Markovian component plus a Martingale-difference component. When the Martingale-difference noise is bounded, we show that the tail of the error can be sub-Gaussian, sub-Weibull, or something lighter than any Pareto but heavier than any Weibull, depending on the step size sequence and on whether the random operator is almost surely contractive, almost surely non-expansive, or expansive with positive probability. Our analysis relies on a novel Lyapunov function involving the moment-generating function of the solution to a Poisson equation, together with an auxiliary projected algorithm. We complement the upper bounds with worst-case examples showing that qualitatively sharper bounds are impossible. We further study the case of unbounded Martingale-difference noise when the average operator is contractive, and the step sizes are of order $1/k$. In this setting, we show that if the random operator is almost surely non-expansive, then the error tail is at most three times heavier than the noise tail, whereas if the random operator is expansive with positive probability, then the error may have substantially heavier tails. These results are obtained through a novel black-box truncation argument that reduces the unbounded-noise setting to the bounded-noise case.

AlltoAll dispatch is the dominant bottleneck of MoE expert parallelism, and the interconnect community has responded with four families of mitigations: predictive sample placement, adaptive expert relayout, hierarchical collectives, and EP-aware topology. All four rest on two assumptions about the workload. The first is that routing imbalance is correctable by the system layer. The second is that the mock-token benchmarks evaluating them faithfully represent production routing. We introduce DODOCO to test both assumptions. We instrument five MoE checkpoints spanning five sequence-mixer designs (DeepSeek-V2-Lite MLA, DeepSeek-MoE-16B MHA, Qwen3-30B GQA, Nemotron-30B Mamba-2, Qwen3.5-35B GDN) under a 5 by 6 grid of data conditions plus a matched EP scan from 4 to 32 ranks on H100s; both assumptions fail. Scaling EP changes the per-expert max/mean token ratio by at most 5% within every architecture's measurable range: the straggler is intrinsic to the routing decision the model makes, not to how its experts land on ranks. Mock tokens overestimate routing Gini by up to a factor of 2.35 and fabricate a batch-size scaling trend that vanishes the moment real text replaces random IDs. A third pattern, unexpected, emerges from the same matrix: the five architectures cleave into two stable bands. MHA and Mamba-2 (data-resilient) drop to Gini 0.105 and 0.150 on wikitext. MLA and GDN (persistently concentrated) stay above 0.24 on every real-text condition and reach 0.29 to 0.38 on mock. GQA is the intermediate case. These bands, not the EP degree or the mock-data profile, are the right workload input to AlltoAll-aware interconnect and dispatch design.

Distributed LLM agent workflows should not be monitored as if they produced a single sequential log. In an asynchronous execution, a decision can only depend on events that are causally visible to the lifeline that makes it: an event that appears earlier in some log may still be unknown locally. We extend the ZipperGen agent-workflow framework with Causal Past Logic (CPL), a small past-time temporal logic for guards in conditionals and while loops. In addition to standard past-time modalities such as previous and since, a guard can inspect the latest causally visible event of another lifeline and selected variables stored there. The formula is a source-level guard: it is evaluated online by the owner lifeline and can influence control flow at runtime. We give a vector-clock monitor with latest-value views and prove that the locally computed monitor value coincides with the denotational semantics of the guard at the current event. Thus runtime verification becomes part of the coordination language itself, rather than a post-hoc check over an execution log.

Multimodal sarcasm detection requires reasoning over cross-modal incongruities between literal expression and intended meaning, yet the specific analytical perspectives needed vary across samples due to the diversity of sarcastic mechanisms. While recent methods make this analytical process explicit, they still rely on fixed, predefined perspectives that operate independently under hand-crafted routing rules. We argue that multimodal sarcasm detection instead calls for self-elicited multi-perspective reasoning, where a model autonomously generates the perspectives needed for each sample and progressively integrates them into a coherent analysis. To realize this goal, we propose ProCrit, a Proposal-Critic two-agent framework with a proposal agent for multi-perspective reasoning and a critic agent for external evaluation and targeted revision guidance. First, to overcome the lack of process-level supervision in existing sarcasm datasets, ProCrit synthesizes process-level reasoning annotations through a dynamic-role agentic rollout: a strong vision-language model sequentially spawns analytical roles within a shared context, and the resulting multi-role trajectories are flattened into sequences that preserve cross-perspective dependencies while enabling efficient autoregressive generation. Second, to improve reasoning reliability, ProCrit adopts a draft-critique-revise paradigm in which an independent critic identifies reasoning deficiencies and provides targeted natural-language feedback for directed revision. Finally, we develop a mutual-refinement training framework that jointly optimizes proposal drafting and feedback-guided revision via dual-stage reinforcement learning, while refining the critic agent according to the actual effectiveness of its feedback. Experiments on three widely used benchmarks demonstrate the effectiveness of ProCrit.

Reinforcement learning with verifiable rewards (RLVR) has recently unlocked strong reasoning capabilities in large language models (LLMs), triggering rapid exploration of new algorithms and data. However, RLVR training is notoriously inefficient: long-tailed rollouts, tool-induced stalls, and asymmetric resource requirements between rollout and training introduce substantial idle time that cannot be eliminated by job-local optimizations such as synchronous pipelining, asynchronous rollout, or colocated execution. We argue that this inefficiency is structural. While idle gaps are unavoidable within individual RLVR jobs, they are largely anti-correlated across jobs and therefore exploitable at the cluster level. Leveraging this observation, we present PlexRL, a cluster-level runtime for multiplexing unified LLM services across RLVR jobs. By centrally managing model placement, state transitions, and function-level scheduling under strict affinity constraints, PlexRL time-slices LLM execution across jobs to fill otherwise idle periods without expensive model migration. Our implementation and evaluations demonstrate that PlexRL significantly improves effective cluster capacity and reduces user GPU hour cost by maximum 37.58% while preserving algorithmic flexibility and introducing minimal per-job overhead.