120 2026-09-11T20:00:13+00:00 rss-help@arxiv.org python-feedgen cs updates on the arXiv.org e-print archive. oai:arXiv.org:2609.10542v1 2026-09-10T20:00:00+00:00 Walid A. Hanafy, Nader Sehatbakhsh, David Irwin, Mani Srivastava, Prashant Shenoy arXiv:2609.10542v1 Announce Type: new Abstract: The Internet of Things is on a trajectory toward a trillion connected devices deployed across multiple domains. These devices are no longer simple sensing and actuation endpoints; they are mobile platforms with embedded processing and intelligent on-device services. The dominant paradigm of offloading computation to cloud and edge servers is a vertical device-to-server interaction model that cannot scale with this trajectory. What is needed instead is a horizontal paradigm in which devices communicate and collaborate directly, pooling their compute, sensing, and actuation into dynamic, cross-domain clusters rather than offloading to centralized infrastructure. We refer to this paradigm as the Internet of Collaborating Things (IoCT). Realizing this vision requires both a portable and secure execution substrate for heterogeneous hardware and an agentic control plane capable of contextual reasoning, transient trust establishment, and open-world adaptation throughout the device collaboration lifecycle without human intervention. In this position paper, we define this new communication, compute, and collaboration paradigm, articulate the role of agentic edge AI in realizing it, and outline research challenges and future directions toward trustworthy, autonomous IoCT collaboration.

arXiv:2609.10542v1 Announce Type: new Abstract: The Internet of Things is on a trajectory toward a trillion connected devices deployed across multiple domains. These devices are no longer simple sensing and actuation endpoints; they are mobile platforms with embedded processing and intelligent on-device services. The dominant paradigm of offloading computation to cloud and edge servers is a vertical device-to-server interaction model that cannot scale with this trajectory. What is needed instead is a horizontal paradigm in which devices communicate and collaborate directly, pooling their compute, sensing, and actuation into dynamic, cross-domain clusters rather than offloading to centralized infrastructure. We refer to this paradigm as the Internet of Collaborating Things (IoCT). Realizing this vision requires both a portable and secure execution substrate for heterogeneous hardware and an agentic control plane capable of contextual reasoning, transient trust establishment, and open-world adaptation throughout the device collaboration lifecycle without human intervention. In this position paper, we define this new communication, compute, and collaboration paradigm, articulate the role of agentic edge AI in realizing it, and outline research challenges and future directions toward trustworthy, autonomous IoCT collaboration.

2026-09-10T20:00:00+00:00 oai:arXiv.org:2609.10543v1 2026-09-10T20:00:00+00:00 Yuki Nakamura arXiv:2609.10543v1 Announce Type: new Abstract: Privacy-preserving exchange designs price on a coarsened view of order flow. We show that a market maker committed to informationally efficient (posterior-mean) pricing on a signal strictly coarser than the flow it settles necessarily cedes a closed-form welfare transfer to traders -- the privacy subsidy -- and that no rule restricted to the coarse signal is simultaneously efficient and zero-profit against the settled flow. We establish this impossibility for a general coarsening, then characterise the subsidy in closed form across three canonical microstructure models: single-period Kyle with Gaussian flow noise, Glosten-Milgrom with a binary direction channel, and continuous-time Kyle-Back with a Brownian channel. The subsidy obeys a structural correspondence with Loss-Versus-Rebalancing, both welfare rates factorising as a squared noise driver times a committed-object factor. Gross of fees the subsidy is a pure transfer recovered by a break-even fee; once levied, that fee distorts volume, and the resulting deadweight -- under an explicit allocative value of trade -- is fourth-order in the noise scale while the gross subsidy is second-order, so privacy is welfare-neutral to leading order with a strictly smaller irrecoverable loss. Endogenising the privacy level, a protocol trading a differential-privacy benefit against this deadweight chooses an interior noise scale in closed form; doing so leaves the half-revealing product of price impact and informed intensity intact while unpinning the volatility-elasticity of price impact from its textbook value of one.

arXiv:2609.10543v1 Announce Type: new Abstract: Privacy-preserving exchange designs price on a coarsened view of order flow. We show that a market maker committed to informationally efficient (posterior-mean) pricing on a signal strictly coarser than the flow it settles necessarily cedes a closed-form welfare transfer to traders -- the privacy subsidy -- and that no rule restricted to the coarse signal is simultaneously efficient and zero-profit against the settled flow. We establish this impossibility for a general coarsening, then characterise the subsidy in closed form across three canonical microstructure models: single-period Kyle with Gaussian flow noise, Glosten-Milgrom with a binary direction channel, and continuous-time Kyle-Back with a Brownian channel. The subsidy obeys a structural correspondence with Loss-Versus-Rebalancing, both welfare rates factorising as a squared noise driver times a committed-object factor. Gross of fees the subsidy is a pure transfer recovered by a break-even fee; once levied, that fee distorts volume, and the resulting deadweight -- under an explicit allocative value of trade -- is fourth-order in the noise scale while the gross subsidy is second-order, so privacy is welfare-neutral to leading order with a strictly smaller irrecoverable loss. Endogenising the privacy level, a protocol trading a differential-privacy benefit against this deadweight chooses an interior noise scale in closed form; doing so leaves the half-revealing product of price impact and informed intensity intact while unpinning the volatility-elasticity of price impact from its textbook value of one.

2026-09-10T20:00:00+00:00 oai:arXiv.org:2609.10545v1 2026-09-10T20:00:00+00:00 Akshay Mhatre, Vikram Karthick, Deepti Gupta, Jia Zou arXiv:2609.10545v1 Announce Type: new Abstract: Federated Learning (FL) enables distributed model training without centralizing data but suffers from high communication overhead. To address this, we propose Block-Wise Marginal Utility Contribution (MUC), a framework that selectively transmits only the most impactful data blocks based on their contribution to model performance. To evaluate our framework, we apply it to a multimodal dataset integrated from multiple MIMIC clinical datasets and show that only 24 out of 1,135 candidate blocks (1.76%) carry meaningful improvement signals, enabling a potential communication reduction of 45-50% while maintaining or improving model quality. Our deduplication-based block selection achieves a macro F1 score of 0.8566 compared to 0.8155 for standard federated optimization, demonstrating that selective transmission can improve performance, particularly in underrepresented classes.

arXiv:2609.10545v1 Announce Type: new Abstract: Federated Learning (FL) enables distributed model training without centralizing data but suffers from high communication overhead. To address this, we propose Block-Wise Marginal Utility Contribution (MUC), a framework that selectively transmits only the most impactful data blocks based on their contribution to model performance. To evaluate our framework, we apply it to a multimodal dataset integrated from multiple MIMIC clinical datasets and show that only 24 out of 1,135 candidate blocks (1.76%) carry meaningful improvement signals, enabling a potential communication reduction of 45-50% while maintaining or improving model quality. Our deduplication-based block selection achieves a macro F1 score of 0.8566 compared to 0.8155 for standard federated optimization, demonstrating that selective transmission can improve performance, particularly in underrepresented classes.

2026-09-10T20:00:00+00:00 oai:arXiv.org:2609.10546v1 2026-09-10T20:00:00+00:00 Juan M. C. Larrosa, Fernando Tohm\'e arXiv:2609.10546v1 Announce Type: new Abstract: This paper is an extension of the approach of Larrosa and Tohm\'e (2003), in which the payoff function is modified by allowing information to flow in both directions. Costs continue to be paid by the agent who initiates the connection, so this asymmetry reveals changes in the final equilibrium topology. We find that several optimal topologies persist as Nash networks, but that a strict Nash network corresponds to the sequential line network with intermediate active nodes; that is, agents are placed in a line and, every other agent, they connect with the predecessor agent and with the subsequent agent. In this way, cost accumulation is interrupted and benefits are maximized.

arXiv:2609.10546v1 Announce Type: new Abstract: This paper is an extension of the approach of Larrosa and Tohm\'e (2003), in which the payoff function is modified by allowing information to flow in both directions. Costs continue to be paid by the agent who initiates the connection, so this asymmetry reveals changes in the final equilibrium topology. We find that several optimal topologies persist as Nash networks, but that a strict Nash network corresponds to the sequential line network with intermediate active nodes; that is, agents are placed in a line and, every other agent, they connect with the predecessor agent and with the subsequent agent. In this way, cost accumulation is interrupted and benefits are maximized.

2026-09-10T20:00:00+00:00 oai:arXiv.org:2609.10547v1 2026-09-10T20:00:00+00:00 John deVadoss arXiv:2609.10547v1 Announce Type: new Abstract: Digital-token standards such as ERC-20, ERC-721, and ERC-1155 have been essential to blockchain adoption because they standardize callable interfaces. Interface standardization, however, does not fully specify token meaning. Two implementations may expose the same transfer function while encoding different assumptions about supply, divisibility, redemption, cancellation, evidence, governance, and lifecycle finality; conversely, two semantically equivalent tokens may be implemented on different ledgers and through different transaction models. This paper presents the InterWork Alliance Token Taxonomy Framework (TTF) as a typed semantic composition model for implementation neutral token specifications. We formalize TTF artifacts, token formulas, behavior and property-set composition, well-formedness constraints, and a layered conformance model covering formula, artifact, message, state, and trace conformance. We further define a reference validation procedure and show how platform neutral control messages can induce conformance obligations and implementation tests. The model is evaluated analytically through document-token, warehouse-receipt, and carbon/digital-MRV case studies, together with representative invalid compositions that interface standards alone do not expose. The analysis shows that token semantics can be specified, compared, validated, mapped, and governed independently of platform binding, providing a foundation for more reliable token interoperability across smart contract platforms, permissioned ledgers, and shared-state systems.

arXiv:2609.10547v1 Announce Type: new Abstract: Digital-token standards such as ERC-20, ERC-721, and ERC-1155 have been essential to blockchain adoption because they standardize callable interfaces. Interface standardization, however, does not fully specify token meaning. Two implementations may expose the same transfer function while encoding different assumptions about supply, divisibility, redemption, cancellation, evidence, governance, and lifecycle finality; conversely, two semantically equivalent tokens may be implemented on different ledgers and through different transaction models. This paper presents the InterWork Alliance Token Taxonomy Framework (TTF) as a typed semantic composition model for implementation neutral token specifications. We formalize TTF artifacts, token formulas, behavior and property-set composition, well-formedness constraints, and a layered conformance model covering formula, artifact, message, state, and trace conformance. We further define a reference validation procedure and show how platform neutral control messages can induce conformance obligations and implementation tests. The model is evaluated analytically through document-token, warehouse-receipt, and carbon/digital-MRV case studies, together with representative invalid compositions that interface standards alone do not expose. The analysis shows that token semantics can be specified, compared, validated, mapped, and governed independently of platform binding, providing a foundation for more reliable token interoperability across smart contract platforms, permissioned ledgers, and shared-state systems.

2026-09-10T20:00:00+00:00 oai:arXiv.org:2609.10548v1 2026-09-10T20:00:00+00:00 Wenji Bai, Muhammad Waseem, Zeeshan Rasheed, Jaakko Peltonen, Pekka Abrahamsson arXiv:2609.10548v1 Announce Type: new Abstract: LLM-based agents for automated code repair have received significant attention in recent years from both research and software engineering practice perspectives. However, limited attention has been paid to patches that pass syntactic and functional verification but still retain or introduce security vulnerabilities. The aim of this research is to systematically identify and categorize such silent failures in LLM-based agentic code repair. We conducted an empirical study using 1,030 valid execution traces produced by seven agent frameworks with GPT-4o-mini across two security-focused datasets, SecurityEval and CVEfixes. Through three iterations of qualitative coding and manual verification, 170 confirmed silent failures were identified. The key results are: (i) Three main categories of silent failures were identified: Omission, Introduction, and Inadequacy. Omission accounts for 48.2% of the confirmed failures, Introduction for 30.6%, and Inadequacy for 21.2%. (ii) Ten fine-grained failure codes were classified under these three categories, showing how agents omit required security controls, apply incomplete defenses, or introduce new vulnerabilities during repair. (iii) Current test-passing evaluation and LLM-based reviewer roles were insufficient to expose or intercept these failures in the confirmed cases. (iv) Similar insecure solutions appeared across different frameworks, suggesting possible shared model-, prompt-, or task-level influences, while single-agent and multi-agent systems showed different failure profiles. The results of this study will assist researchers and practitioners in improving the evaluation of LLM-based agentic code repair and developing targeted verification methods that go beyond functional correctness and cover all generated artifacts.

arXiv:2609.10548v1 Announce Type: new Abstract: LLM-based agents for automated code repair have received significant attention in recent years from both research and software engineering practice perspectives. However, limited attention has been paid to patches that pass syntactic and functional verification but still retain or introduce security vulnerabilities. The aim of this research is to systematically identify and categorize such silent failures in LLM-based agentic code repair. We conducted an empirical study using 1,030 valid execution traces produced by seven agent frameworks with GPT-4o-mini across two security-focused datasets, SecurityEval and CVEfixes. Through three iterations of qualitative coding and manual verification, 170 confirmed silent failures were identified. The key results are: (i) Three main categories of silent failures were identified: Omission, Introduction, and Inadequacy. Omission accounts for 48.2% of the confirmed failures, Introduction for 30.6%, and Inadequacy for 21.2%. (ii) Ten fine-grained failure codes were classified under these three categories, showing how agents omit required security controls, apply incomplete defenses, or introduce new vulnerabilities during repair. (iii) Current test-passing evaluation and LLM-based reviewer roles were insufficient to expose or intercept these failures in the confirmed cases. (iv) Similar insecure solutions appeared across different frameworks, suggesting possible shared model-, prompt-, or task-level influences, while single-agent and multi-agent systems showed different failure profiles. The results of this study will assist researchers and practitioners in improving the evaluation of LLM-based agentic code repair and developing targeted verification methods that go beyond functional correctness and cover all generated artifacts.

2026-09-10T20:00:00+00:00 oai:arXiv.org:2609.10549v1 2026-09-10T20:00:00+00:00 Guangyu Xiang, Lin Zhang, Haoxuan Yu, Xinglin Pan, Shaohuai Shi, Xiaowen Chu arXiv:2609.10549v1 Announce Type: new Abstract: Overlapping communication and computation operators is a common practice to hide communication overheads, accelerating large language models (LLMs) training on GPU clusters. Existing systems achieve this through either intra-operator fusion (IntraFusion), which packs operators into a single large kernel, or inter-operator decomposition (InterDecom), which splits a tensor into multiple parts for pipelined execution. However, current IntraFusion methods underutilize network topology, causing suboptimal bandwidth usage on multi-GPU systems, while InterDecom struggles to determine the optimal number of decomposed parts for peak performance. To address these issues, we introduce Compass, which employs systematic optimization and comprehensive modeling. First, we design a novel IntraFusion algorithm leveraging double-ring communications to maximize bandwidth utilization in hybrid NVLink-PCIe systems, achieving 1.5x-2.5x speedups. Second, we develop a decomposition model that mathematically derives the optimal tensor decomposition degree for InterDecom, improving performance by up to 1.3x. Finally, we develop a unified performance framework that accurately determines the best strategy for different scenarios. We validate Compass through extensive evaluation across 288 configurations and end-to-end experiments on real-world applications. The results demonstrate that Compass consistently selects the optimal strategy, achieving up to a 1.42x end-to-end speedup compared to the Megatron-LM baseline.

arXiv:2609.10549v1 Announce Type: new Abstract: Overlapping communication and computation operators is a common practice to hide communication overheads, accelerating large language models (LLMs) training on GPU clusters. Existing systems achieve this through either intra-operator fusion (IntraFusion), which packs operators into a single large kernel, or inter-operator decomposition (InterDecom), which splits a tensor into multiple parts for pipelined execution. However, current IntraFusion methods underutilize network topology, causing suboptimal bandwidth usage on multi-GPU systems, while InterDecom struggles to determine the optimal number of decomposed parts for peak performance. To address these issues, we introduce Compass, which employs systematic optimization and comprehensive modeling. First, we design a novel IntraFusion algorithm leveraging double-ring communications to maximize bandwidth utilization in hybrid NVLink-PCIe systems, achieving 1.5x-2.5x speedups. Second, we develop a decomposition model that mathematically derives the optimal tensor decomposition degree for InterDecom, improving performance by up to 1.3x. Finally, we develop a unified performance framework that accurately determines the best strategy for different scenarios. We validate Compass through extensive evaluation across 288 configurations and end-to-end experiments on real-world applications. The results demonstrate that Compass consistently selects the optimal strategy, achieving up to a 1.42x end-to-end speedup compared to the Megatron-LM baseline.

2026-09-10T20:00:00+00:00 oai:arXiv.org:2609.10550v1 2026-09-10T20:00:00+00:00 Tejinder Singh, John Pflueger, Jeebak Mitra, Robert Lincourt, Mitchell Markow, Bhavesh A. Patel arXiv:2609.10550v1 Announce Type: new Abstract: AI deployment performance is shaped not by model architecture alone, but by interactions among compression, compiler transformations, and serving policies. Published benchmarks often report latency and throughput under incomparable conditions, limiting their use for deployment decisions. This paper presents a unified analytical treatment of inference optimization across the deployment stack. We introduce a three-layer taxonomy covering model-level techniques such as quantization, pruning, and distillation; compiler transformations such as graph fusion, layout optimization, and kernel autotuning; and system policies such as dynamic batching, admission control, and memory tiering. We formulate deployment as a constrained multi-objective optimization problem over accuracy, latency, throughput, memory footprint, and energy, and analyze a deployment-ranking functional with Pareto monotonicity and scale invariance. Roofline models show how memory-bandwidth hierarchies bound performance across precision regimes, while queuing models explain how service-time changes amplify response time under load. To improve comparability, we propose an evidence protocol that separates measured, derived, and analytical claims; limits numerical comparison to within-paper results; and requires reporting of hardware, software versions, batch semantics, and thermal state. We synthesize evidence from edge platforms, including Jetson AGX Orin and five inference frameworks; data center GPUs, including A100 and H100 with three LLM serving engines; and quantization studies across the Llama-3.1 family. The synthesis shows that deployment outcomes are governed by cross-layer interactions that no single-layer analysis can predict. We conclude with a constraint-aware selection procedure and open problems in compiler-serving co-optimization, cross-hardware performance prediction, and standardized energy reporting.

arXiv:2609.10550v1 Announce Type: new Abstract: AI deployment performance is shaped not by model architecture alone, but by interactions among compression, compiler transformations, and serving policies. Published benchmarks often report latency and throughput under incomparable conditions, limiting their use for deployment decisions. This paper presents a unified analytical treatment of inference optimization across the deployment stack. We introduce a three-layer taxonomy covering model-level techniques such as quantization, pruning, and distillation; compiler transformations such as graph fusion, layout optimization, and kernel autotuning; and system policies such as dynamic batching, admission control, and memory tiering. We formulate deployment as a constrained multi-objective optimization problem over accuracy, latency, throughput, memory footprint, and energy, and analyze a deployment-ranking functional with Pareto monotonicity and scale invariance. Roofline models show how memory-bandwidth hierarchies bound performance across precision regimes, while queuing models explain how service-time changes amplify response time under load. To improve comparability, we propose an evidence protocol that separates measured, derived, and analytical claims; limits numerical comparison to within-paper results; and requires reporting of hardware, software versions, batch semantics, and thermal state. We synthesize evidence from edge platforms, including Jetson AGX Orin and five inference frameworks; data center GPUs, including A100 and H100 with three LLM serving engines; and quantization studies across the Llama-3.1 family. The synthesis shows that deployment outcomes are governed by cross-layer interactions that no single-layer analysis can predict. We conclude with a constraint-aware selection procedure and open problems in compiler-serving co-optimization, cross-hardware performance prediction, and standardized energy reporting.

2026-09-10T20:00:00+00:00 oai:arXiv.org:2609.10552v1 2026-09-10T20:00:00+00:00 Florian Atzenhofer-Baumgartner, David Fleischhacker, Max Resch, Lukas Waldhofer, Michael Otto arXiv:2609.10552v1 Announce Type: new Abstract: We describe the design, implementation, and operation of a small federated GPU cluster built for Digital Humanities (DH) research within the Austrian DHinfra.at project. The system spans two university sites, brokers logins from a national identity federation, and exposes compute through three interfaces: interactive notebooks, SSH, and an OpenAI-compatible inference API. The core of the platform is a control plane that maps federated identity, per-project quotas, and a model catalogue onto self-service interfaces. We document this control plane in detail, including the request-and-approval queue that mediates every privileged cluster effect and the two-tier model-serving setup, which combines always-on services with on-demand model swapping and lets researchers consume large language models over HTTPS without holding a cluster account. We report the constraints that shaped the build: a fixed power budget, EU-wide procurement, a team of about 1.5 full-time equivalents, and the relationship between prioritized access and utilization. We also record the main build decisions against the alternatives we researched and did not adopt. The platform is released under the Apache License 2.0. We close with a playbook for teams building a domain-specific GPU cluster on a limited budget. This is a living technical report, versioned and updated as the platform evolves.

arXiv:2609.10552v1 Announce Type: new Abstract: We describe the design, implementation, and operation of a small federated GPU cluster built for Digital Humanities (DH) research within the Austrian DHinfra.at project. The system spans two university sites, brokers logins from a national identity federation, and exposes compute through three interfaces: interactive notebooks, SSH, and an OpenAI-compatible inference API. The core of the platform is a control plane that maps federated identity, per-project quotas, and a model catalogue onto self-service interfaces. We document this control plane in detail, including the request-and-approval queue that mediates every privileged cluster effect and the two-tier model-serving setup, which combines always-on services with on-demand model swapping and lets researchers consume large language models over HTTPS without holding a cluster account. We report the constraints that shaped the build: a fixed power budget, EU-wide procurement, a team of about 1.5 full-time equivalents, and the relationship between prioritized access and utilization. We also record the main build decisions against the alternatives we researched and did not adopt. The platform is released under the Apache License 2.0. We close with a playbook for teams building a domain-specific GPU cluster on a limited budget. This is a living technical report, versioned and updated as the platform evolves.

2026-09-10T20:00:00+00:00 oai:arXiv.org:2609.10554v1 2026-09-10T20:00:00+00:00 Marios Asiminakis, Polydoros Petrakis, Manolis Marazakis arXiv:2609.10554v1 Announce Type: new Abstract: Heterogeneous memory systems that combine high-bandwidth memory (HBM) with commodity DRAM can accelerate bandwidth-bound HPC workloads, but current page placement largely depends on manual tuning or OS heuristics not designed for multi-tier dynamics. We present SHAMBLES, a kernel-integrated framework that profiles application memory behavior at low overhead and migrates data across tiers without requiring application changes. SHAMBLES exposes a policy-agnostic interface and a lightweight user-space runtime with pluggable policies (e.g. recency and frequency based) as well as static placement for controlled studies. A logging mode provides reproducible timelines of allocations and migrations to aid analysis. We implement SHAMBLES on a commodity Linux system with HBM and DDR exposed as NUMA nodes and evaluate it with the HPCG, DGEMM benchmarks and Himeno stencil mini-app. Our design and methodology show how transparent, policy-driven migration can respond to changing access locality and concentrate hot data in HBM without developer intervention, offering a practical path to performance portability on tiered memory. Results from HPCG show that we can maintain up to 93.75% of the all-in-HBM baseline performance, while keeping only 40% of the problem size in the HBM. DGEMM experiments show that dynamic policies in SHAMBLES sustain up to 99% of the all-in-HBM performance, while keeping only one third of the DGEMM matrix footprint in HBM. For Himeno, SHAMBLES shows that fast-tier selection must be both workload-aware and size-aware: with a 50% fast-tier budget, it can outperform fixed all-in-HBM and all-in-DDR placements for the L size, while the XL size shifts back toward HBM.

arXiv:2609.10554v1 Announce Type: new Abstract: Heterogeneous memory systems that combine high-bandwidth memory (HBM) with commodity DRAM can accelerate bandwidth-bound HPC workloads, but current page placement largely depends on manual tuning or OS heuristics not designed for multi-tier dynamics. We present SHAMBLES, a kernel-integrated framework that profiles application memory behavior at low overhead and migrates data across tiers without requiring application changes. SHAMBLES exposes a policy-agnostic interface and a lightweight user-space runtime with pluggable policies (e.g. recency and frequency based) as well as static placement for controlled studies. A logging mode provides reproducible timelines of allocations and migrations to aid analysis. We implement SHAMBLES on a commodity Linux system with HBM and DDR exposed as NUMA nodes and evaluate it with the HPCG, DGEMM benchmarks and Himeno stencil mini-app. Our design and methodology show how transparent, policy-driven migration can respond to changing access locality and concentrate hot data in HBM without developer intervention, offering a practical path to performance portability on tiered memory. Results from HPCG show that we can maintain up to 93.75% of the all-in-HBM baseline performance, while keeping only 40% of the problem size in the HBM. DGEMM experiments show that dynamic policies in SHAMBLES sustain up to 99% of the all-in-HBM performance, while keeping only one third of the DGEMM matrix footprint in HBM. For Himeno, SHAMBLES shows that fast-tier selection must be both workload-aware and size-aware: with a 50% fast-tier budget, it can outperform fixed all-in-HBM and all-in-DDR placements for the L size, while the XL size shifts back toward HBM.

2026-09-10T20:00:00+00:00 oai:arXiv.org:2609.10556v1 2026-09-10T20:00:00+00:00 Sukanya Samanta, Kei Kimura, Makoto Yokoo, Palash Dey arXiv:2609.10556v1 Announce Type: new Abstract: Interdicting an escaping criminal with limited police resources on large-scale transportation networks is a challenging problem due to the dynamic nature of both attacker movement and defender deployment. This paper proposes \emph{MLN-EIGS}, a multilayer network-based framework for solving dynamic escape interdiction problems formulated as a Stackelberg security game. A time-expanded multilayer network is constructed to explicitly model the temporal evolution of the transportation network and the feasible movements of both the attacker and the defenders. The attacker seeks to maximize the probability of successful escape, while the defenders aim to maximize the probability of interdiction. To efficiently compute the attacker's best response, the probabilistic escape formulation is transformed into an equivalent shortest-path problem through a logarithmic transformation, enabling the use of Dijkstra's algorithm. Since the defender best-response problem is computationally intractable, an approximation defender oracle is developed to generate high-quality defender strategies on the multilayer network. The proposed MLN-EIGS framework is benchmarked against an exact mixed-integer linear programming (MILP)-based Stackelberg formulation on a large real-world transportation network. Computational experiments demonstrate that MLN-EIGS consistently achieves defender utilities that closely match those of the exact MILP approach while substantially reducing computational time. These results demonstrate that the proposed MLN-EIGS framework provides an effective, computationally efficient, and scalable alternative to exact MILP-based Stackelberg optimization for large-scale dynamic escape interdiction problems.

arXiv:2609.10556v1 Announce Type: new Abstract: Interdicting an escaping criminal with limited police resources on large-scale transportation networks is a challenging problem due to the dynamic nature of both attacker movement and defender deployment. This paper proposes \emph{MLN-EIGS}, a multilayer network-based framework for solving dynamic escape interdiction problems formulated as a Stackelberg security game. A time-expanded multilayer network is constructed to explicitly model the temporal evolution of the transportation network and the feasible movements of both the attacker and the defenders. The attacker seeks to maximize the probability of successful escape, while the defenders aim to maximize the probability of interdiction. To efficiently compute the attacker's best response, the probabilistic escape formulation is transformed into an equivalent shortest-path problem through a logarithmic transformation, enabling the use of Dijkstra's algorithm. Since the defender best-response problem is computationally intractable, an approximation defender oracle is developed to generate high-quality defender strategies on the multilayer network. The proposed MLN-EIGS framework is benchmarked against an exact mixed-integer linear programming (MILP)-based Stackelberg formulation on a large real-world transportation network. Computational experiments demonstrate that MLN-EIGS consistently achieves defender utilities that closely match those of the exact MILP approach while substantially reducing computational time. These results demonstrate that the proposed MLN-EIGS framework provides an effective, computationally efficient, and scalable alternative to exact MILP-based Stackelberg optimization for large-scale dynamic escape interdiction problems.

2026-09-10T20:00:00+00:00 oai:arXiv.org:2609.10557v1 2026-09-10T20:00:00+00:00 Berry Gerrits, Fabian Akkerman arXiv:2609.10557v1 Announce Type: new Abstract: Customer choice models are central to revenue management in attended home delivery, yet the field usually picks one without much thought. Does the choice model matter? We believe it does, and this paper shows that revenue management policies judged by their own assumed model, or trusted under a single set of parameters, can be worse than they look.

arXiv:2609.10557v1 Announce Type: new Abstract: Customer choice models are central to revenue management in attended home delivery, yet the field usually picks one without much thought. Does the choice model matter? We believe it does, and this paper shows that revenue management policies judged by their own assumed model, or trusted under a single set of parameters, can be worse than they look.

2026-09-10T20:00:00+00:00 oai:arXiv.org:2609.10559v1 2026-09-10T20:00:00+00:00 Wenzhe Jin, Haina Tang arXiv:2609.10559v1 Announce Type: new Abstract: To address the challenges of behavioral multimodality, limited semantic utilization, and long-term error accumulation in vessel trajectory prediction, this paper proposes M3-Former, a multimodal trajectory prediction framework enhanced by large language models (LLMs). The proposed framework incorporates vessel static attributes and navigational intent as semantic priors for long-term trajectory modeling. Specifically, a unified multimodal representation space is constructed, in which static semantic information is encoded by a pre-trained LLM and aligned with dynamic trajectory features through self-attention. To jointly capture global route planning and local motion variations, a dual-granularity Mixture-of-Experts (MoE) architecture is introduced, where sequence-level experts model global navigation trends and token-level experts refine fine-grained maneuvering behaviors. In addition, a Steering-Weighted Cross-Entropy loss is designed to alleviate the long-tail distribution of sparse turning samples and improve prediction accuracy in critical maneuvering scenarios. Experiments on a real-world Danish AIS dataset demonstrate that M\textsuperscript{3}-Former consistently outperforms state-of-the-art baselines across prediction horizons from 1 to 4 hours. In the 4-hour prediction task, the proposed method reduces Average Displacement Error (ADE) and Final Displacement Error (FDE) by 4.4\% and 5.1\%, respectively, compared with the strongest baseline. Qualitative and ablation analyses further verify that semantic fusion effectively reduces long-term trajectory drift, while the dual-granularity MoE improves robustness in complex waterways and route-branching scenarios. The proposed framework establishes a semantic-guided hierarchical prediction paradigm, in which high-level navigational intent and local motion dynamics are jointly modeled for robust long-term vessel trajectory forecasting.

arXiv:2609.10559v1 Announce Type: new Abstract: To address the challenges of behavioral multimodality, limited semantic utilization, and long-term error accumulation in vessel trajectory prediction, this paper proposes M3-Former, a multimodal trajectory prediction framework enhanced by large language models (LLMs). The proposed framework incorporates vessel static attributes and navigational intent as semantic priors for long-term trajectory modeling. Specifically, a unified multimodal representation space is constructed, in which static semantic information is encoded by a pre-trained LLM and aligned with dynamic trajectory features through self-attention. To jointly capture global route planning and local motion variations, a dual-granularity Mixture-of-Experts (MoE) architecture is introduced, where sequence-level experts model global navigation trends and token-level experts refine fine-grained maneuvering behaviors. In addition, a Steering-Weighted Cross-Entropy loss is designed to alleviate the long-tail distribution of sparse turning samples and improve prediction accuracy in critical maneuvering scenarios. Experiments on a real-world Danish AIS dataset demonstrate that M\textsuperscript{3}-Former consistently outperforms state-of-the-art baselines across prediction horizons from 1 to 4 hours. In the 4-hour prediction task, the proposed method reduces Average Displacement Error (ADE) and Final Displacement Error (FDE) by 4.4\% and 5.1\%, respectively, compared with the strongest baseline. Qualitative and ablation analyses further verify that semantic fusion effectively reduces long-term trajectory drift, while the dual-granularity MoE improves robustness in complex waterways and route-branching scenarios. The proposed framework establishes a semantic-guided hierarchical prediction paradigm, in which high-level navigational intent and local motion dynamics are jointly modeled for robust long-term vessel trajectory forecasting.

2026-09-10T20:00:00+00:00 oai:arXiv.org:2609.10565v1 2026-09-10T20:00:00+00:00 Marco Tulio Angulo arXiv:2609.10565v1 Announce Type: new Abstract: Connectivity can turn separate contributions into a public good that benefits participants and nonparticipants alike. Yet only participants pay for it, and each can avoid her charge by withdrawing while enjoying whatever public output survives. How do network connections constrain who can or must pay, and how large a cost can the group bear? We study a participation game in which public output equals the size of the largest connected group of active players, whose members divide a fixed operating cost. The output a participant destroys by withdrawing is her withdrawal responsibility and bounds her payment. These bounds determine the group's cost-bearing capacity and, whenever the group can be kept active, which participants must pay. We prove that among connected networks with at least five players, the path uniquely maximizes the cost-bearing capacity of a group, yet minimizes the expected surviving output after a uniformly selected participant withdraws. As the group grows, the ratio of maximal capacity to gross social benefit under full participation converges to one quarter. Fragility can therefore strengthen voluntary finance, but even the most financeable network cannot close the gap between the value it creates and what its participants will pay to sustain it.

arXiv:2609.10565v1 Announce Type: new Abstract: Connectivity can turn separate contributions into a public good that benefits participants and nonparticipants alike. Yet only participants pay for it, and each can avoid her charge by withdrawing while enjoying whatever public output survives. How do network connections constrain who can or must pay, and how large a cost can the group bear? We study a participation game in which public output equals the size of the largest connected group of active players, whose members divide a fixed operating cost. The output a participant destroys by withdrawing is her withdrawal responsibility and bounds her payment. These bounds determine the group's cost-bearing capacity and, whenever the group can be kept active, which participants must pay. We prove that among connected networks with at least five players, the path uniquely maximizes the cost-bearing capacity of a group, yet minimizes the expected surviving output after a uniformly selected participant withdraws. As the group grows, the ratio of maximal capacity to gross social benefit under full participation converges to one quarter. Fragility can therefore strengthen voluntary finance, but even the most financeable network cannot close the gap between the value it creates and what its participants will pay to sustain it.

2026-09-10T20:00:00+00:00 oai:arXiv.org:2609.10572v1 2026-09-10T20:00:00+00:00 Ranjit Raut, Aarav Subedi, Ashim Shrestha arXiv:2609.10572v1 Announce Type: new Abstract: Non-Latin handwritten character recognition (HCR) remains understudied. Dominant methods consider it as generic image classification, which uses model scale to implicitly learn stroke structure. Structural-prior efficiency---the principle that explicitly encoding script-geometric regularities as architectural inductive biases can be both more accurate and require fewer parameters. We introduce GraphemeNet, a unified multi-script architecture, governed by two orthogonal binary axes. Axis 1 operationalises stroke-level geometric regularity via Persistent Scaffold Injection (PSI): a script-specific asymmetric convolution injects a stroke scaffold as a weighted residual at every encoder stage, continuously anchoring learned features to script geometry---distinct from skip connections, auxiliary losses, or attention reweighting. Axis 2 selects between global average pooling with gated fusion and cross-scale attention with a Stroke Topology Module (STM), depending on whether glyph discrimination requires spatial relational reasoning. A Linear Capsule Routing (LCR) with $O(n)$ routing is shared universally. On fourteen benchmarks across eight writing systems, the architecture generalises with only scaffold and decoder topology varying per script, consistently challenging, outperforming published baselines, and establishing structural-prior efficiency as a broadly applicable principle for multi-script HCR.

arXiv:2609.10572v1 Announce Type: new Abstract: Non-Latin handwritten character recognition (HCR) remains understudied. Dominant methods consider it as generic image classification, which uses model scale to implicitly learn stroke structure. Structural-prior efficiency---the principle that explicitly encoding script-geometric regularities as architectural inductive biases can be both more accurate and require fewer parameters. We introduce GraphemeNet, a unified multi-script architecture, governed by two orthogonal binary axes. Axis 1 operationalises stroke-level geometric regularity via Persistent Scaffold Injection (PSI): a script-specific asymmetric convolution injects a stroke scaffold as a weighted residual at every encoder stage, continuously anchoring learned features to script geometry---distinct from skip connections, auxiliary losses, or attention reweighting. Axis 2 selects between global average pooling with gated fusion and cross-scale attention with a Stroke Topology Module (STM), depending on whether glyph discrimination requires spatial relational reasoning. A Linear Capsule Routing (LCR) with $O(n)$ routing is shared universally. On fourteen benchmarks across eight writing systems, the architecture generalises with only scaffold and decoder topology varying per script, consistently challenging, outperforming published baselines, and establishing structural-prior efficiency as a broadly applicable principle for multi-script HCR.

2026-09-10T20:00:00+00:00 oai:arXiv.org:2609.10576v1 2026-09-10T20:00:00+00:00 Emirhan Tekez, Melih Bastopcu, Sinan Gezici arXiv:2609.10576v1 Announce Type: new Abstract: We consider a fully connected gossip network of $n$ nodes that track a binary continuous-time Markov source through a strategic sender transmitting updates under a communication budget at a rate that depends on the source state. The receivers exchange packets through gossip and decide whether to follow the sender. We model this interaction as a Stackelberg game and analyze it through a stochastic hybrid systems (SHS) framework. We prove that the sender's budget constraint binds at every interior optimum, reducing its problem to a one-dimensional search on the budget line. When the sender pushes its preferred state at the higher rate, the receivers gossip at the highest available rate. Gossip has no direction of its own and works against the asymmetry in the sender's policy rather than reinforcing it. We prove that an optimistic Stackelberg equilibrium exists, and that it is unique and explicitly characterized whenever a policy on the strategic half of that line is feasible at the gossip cap. Monte Carlo simulations agree with the analytical recursion.

arXiv:2609.10576v1 Announce Type: new Abstract: We consider a fully connected gossip network of $n$ nodes that track a binary continuous-time Markov source through a strategic sender transmitting updates under a communication budget at a rate that depends on the source state. The receivers exchange packets through gossip and decide whether to follow the sender. We model this interaction as a Stackelberg game and analyze it through a stochastic hybrid systems (SHS) framework. We prove that the sender's budget constraint binds at every interior optimum, reducing its problem to a one-dimensional search on the budget line. When the sender pushes its preferred state at the higher rate, the receivers gossip at the highest available rate. Gossip has no direction of its own and works against the asymmetry in the sender's policy rather than reinforcing it. We prove that an optimistic Stackelberg equilibrium exists, and that it is unique and explicitly characterized whenever a policy on the strategic half of that line is feasible at the gossip cap. Monte Carlo simulations agree with the analytical recursion.

2026-09-10T20:00:00+00:00 oai:arXiv.org:2609.10580v1 2026-09-10T20:00:00+00:00 Moses Boudourides arXiv:2609.10580v1 Announce Type: new Abstract: In a mechanistic model of the dawn chorus, Kaye showed that heterogeneous activation thresholds and a shared feedback signal determined by the population's active fraction can produce abrupt collective activation and hysteresis. We extend this mechanism to a network of interacting agents. Each node has a continuous activation level, and a nonnegative row-stochastic matrix determines how node activities contribute to individual feedback. We prove that sufficiently weak feedback yields a unique globally attracting equilibrium. For all feedback strengths, the homogeneous dynamics exactly reproduce Kaye's scalar equation; consequently, network topology does not alter the folds or cusp of the homogeneous branch, and no heterogeneous mode becomes unstable before the homogeneous mode. For equitable partitions, the network admits an exact quotient system in which nodes within a block receive the same aggregate input from every block. When blocks are uncoupled, the quotient reduces to independent copies of Kaye's scalar equation. We show that every assignment of stable scalar equilibria to blocks persists under sufficiently weak interblock coupling, producing a combinatorial family of stable quotient equilibria that lift to stable full-network equilibria and remain under small perturbations that break exact equitability. For two symmetrically coupled blocks, branches with unequal block activities terminate at a pair of symmetry-related cusp bifurcations. Near the onset of bistability, we derive scaling laws for the interblock coupling at which these bifurcations occur, the activity difference between the blocks at bifurcation, and the corresponding shift of the external stimulus from the scalar cusp. Numerical continuation confirms the scaling laws for gamma, logistic, and normal threshold distributions.

arXiv:2609.10580v1 Announce Type: new Abstract: In a mechanistic model of the dawn chorus, Kaye showed that heterogeneous activation thresholds and a shared feedback signal determined by the population's active fraction can produce abrupt collective activation and hysteresis. We extend this mechanism to a network of interacting agents. Each node has a continuous activation level, and a nonnegative row-stochastic matrix determines how node activities contribute to individual feedback. We prove that sufficiently weak feedback yields a unique globally attracting equilibrium. For all feedback strengths, the homogeneous dynamics exactly reproduce Kaye's scalar equation; consequently, network topology does not alter the folds or cusp of the homogeneous branch, and no heterogeneous mode becomes unstable before the homogeneous mode. For equitable partitions, the network admits an exact quotient system in which nodes within a block receive the same aggregate input from every block. When blocks are uncoupled, the quotient reduces to independent copies of Kaye's scalar equation. We show that every assignment of stable scalar equilibria to blocks persists under sufficiently weak interblock coupling, producing a combinatorial family of stable quotient equilibria that lift to stable full-network equilibria and remain under small perturbations that break exact equitability. For two symmetrically coupled blocks, branches with unequal block activities terminate at a pair of symmetry-related cusp bifurcations. Near the onset of bistability, we derive scaling laws for the interblock coupling at which these bifurcations occur, the activity difference between the blocks at bifurcation, and the corresponding shift of the external stimulus from the scalar cusp. Numerical continuation confirms the scaling laws for gamma, logistic, and normal threshold distributions.

2026-09-10T20:00:00+00:00 oai:arXiv.org:2609.10584v1 2026-09-10T20:00:00+00:00 Minh Vu Duc, Trung Le Huu, H\`a Minh Ho\`ang, Trung Thanh Nguyen, Phuong Khanh Nguyen, Huynh Thi Thanh Binh arXiv:2609.10584v1 Announce Type: new Abstract: Bounded-suboptimal search seeks a solution within a factor $w$ of optimal while reducing search effort. Focal Search (FS) uses heuristic guidance within FOCAL, the frontier nodes eligible under the threshold $w f_{\min}$, but its deterministic policy may leave $f_{\min}$ unchanged for many expansions. We introduce Probabilistic Focal Search (PFS), which follows the FS guided choice with probability $p$ and expands a minimum-$f$ OPEN node with probability $1-p$. The latter branch encourages the lower bound to advance, enlarging FOCAL and admitting nodes that may lead to feasible solutions. By balancing guidance and lower-bound advancement, this mechanism can reduce time to a bounded solution when progress is limited by delayed FOCAL admission. As a secondary transfer experiment, we apply the same scheduler to Dynamic Potential Search, yielding Probabilistic Dynamic Potential Search (PDPS). We benchmark PFS against FS on N-Puzzle, Pancake Sorting, and the Traveling Salesperson Problem (TSP), and evaluate its anytime extension on the Generalized Covering TSP (GCTSP), using multiple $w$ and $p$ values. Across these benchmarks, the largest gains occur when long $f_{\min}$ plateaus delay useful FOCAL admissions; in such settings, the probabilistic factor may reduce node expansions by about 90\% or more (e.g., on N-Puzzle and TSP). For the anytime algorithm family, Anytime Probabilistic Focal Search (APFS) outperforms all tested algorithms in evaluating anytime methods on GCTSP. We also observe that the benefit is smaller when the deterministic search already advances efficiently (e.g., Pancake Sorting), indicating that the probabilistic factor is most useful when FOCAL admission is a search bottleneck. The PDPS transfer shows that the mechanism also transfers to potential guidance, although its common-success effects remain domain- and bound-dependent.

arXiv:2609.10584v1 Announce Type: new Abstract: Bounded-suboptimal search seeks a solution within a factor $w$ of optimal while reducing search effort. Focal Search (FS) uses heuristic guidance within FOCAL, the frontier nodes eligible under the threshold $w f_{\min}$, but its deterministic policy may leave $f_{\min}$ unchanged for many expansions. We introduce Probabilistic Focal Search (PFS), which follows the FS guided choice with probability $p$ and expands a minimum-$f$ OPEN node with probability $1-p$. The latter branch encourages the lower bound to advance, enlarging FOCAL and admitting nodes that may lead to feasible solutions. By balancing guidance and lower-bound advancement, this mechanism can reduce time to a bounded solution when progress is limited by delayed FOCAL admission. As a secondary transfer experiment, we apply the same scheduler to Dynamic Potential Search, yielding Probabilistic Dynamic Potential Search (PDPS). We benchmark PFS against FS on N-Puzzle, Pancake Sorting, and the Traveling Salesperson Problem (TSP), and evaluate its anytime extension on the Generalized Covering TSP (GCTSP), using multiple $w$ and $p$ values. Across these benchmarks, the largest gains occur when long $f_{\min}$ plateaus delay useful FOCAL admissions; in such settings, the probabilistic factor may reduce node expansions by about 90\% or more (e.g., on N-Puzzle and TSP). For the anytime algorithm family, Anytime Probabilistic Focal Search (APFS) outperforms all tested algorithms in evaluating anytime methods on GCTSP. We also observe that the benefit is smaller when the deterministic search already advances efficiently (e.g., Pancake Sorting), indicating that the probabilistic factor is most useful when FOCAL admission is a search bottleneck. The PDPS transfer shows that the mechanism also transfers to potential guidance, although its common-success effects remain domain- and bound-dependent.

2026-09-10T20:00:00+00:00 oai:arXiv.org:2609.10585v1 2026-09-10T20:00:00+00:00 Xinkai Zhang arXiv:2609.10585v1 Announce Type: new Abstract: We prove that every nonnegative additive chore instance with three agents and either seven or eight indivisible chores admits a chores-EFX allocation, in the zero-tolerant sense that every owned chore, including one of zero cost, is quantified in the trim. Both proofs are computer-assisted, but their machine formulas differ. For seven chores, hand-checkable canonicalization reduces nonexistence to a quantifier-free linear real arithmetic (QF_LRA) formula over 21 variables with one failure clause per complete allocation. For eight chores, instances in which two agents share a weakly cheapest chore are lifted from the seven-chore theorem through the matching insertion lemma of Kobayashi, Mahara, and Sakamoto, and the remaining pairwise-disjoint-argmin class reduces to a residual QF_LRA formula over 24 variables. Z3 5.1.0 and cvc5 1.3.4 report both formulas unsatisfiable. With the known $m\leq 2n$ theorem, this settles every three-agent instance with at most eight chores; $m=9$ is the next open cardinality, and additive chores can fail to admit EFX for every $n\geq 4$.

arXiv:2609.10585v1 Announce Type: new Abstract: We prove that every nonnegative additive chore instance with three agents and either seven or eight indivisible chores admits a chores-EFX allocation, in the zero-tolerant sense that every owned chore, including one of zero cost, is quantified in the trim. Both proofs are computer-assisted, but their machine formulas differ. For seven chores, hand-checkable canonicalization reduces nonexistence to a quantifier-free linear real arithmetic (QF_LRA) formula over 21 variables with one failure clause per complete allocation. For eight chores, instances in which two agents share a weakly cheapest chore are lifted from the seven-chore theorem through the matching insertion lemma of Kobayashi, Mahara, and Sakamoto, and the remaining pairwise-disjoint-argmin class reduces to a residual QF_LRA formula over 24 variables. Z3 5.1.0 and cvc5 1.3.4 report both formulas unsatisfiable. With the known $m\leq 2n$ theorem, this settles every three-agent instance with at most eight chores; $m=9$ is the next open cardinality, and additive chores can fail to admit EFX for every $n\geq 4$.

2026-09-10T20:00:00+00:00 oai:arXiv.org:2609.10588v1 2026-09-10T20:00:00+00:00 Nasser Khalili arXiv:2609.10588v1 Announce Type: new Abstract: This paper formalizes threshold-based selection as an evaluation-gating architecture in which each incumbent is tested before variation and a replacement is generated and evaluated only when contextual pressure exceeds intrinsic strength. The mechanism is instantiated as Leaf Abscission Optimization (LAO), using rank-based strength, a phenological seasonal signal, diversity modulation, environmental pressure, and a base regrowth kernel. A blocked two-to-the-fourth-power factorial analysis at dimension 10 on the CEC 2017 suite reduces the original multi-layer design to a parsimonious core: drift is harmful, while the other three auxiliary layers show no robust independent evidence of benefit. The resulting LAO-Core attains the third-best mean Friedman rank among nine optimizers at dimensions 10, 30, and 50 under the equal evaluation budget. A four-budget sweep shows budget-dependent relative performance, with adaptive differential-evolution baselines gaining relative advantage at larger budgets; the nine-cell dimension-budget analysis establishes neither an evaluation-budget-per-dimension-only law nor a statistically significant dimension-budget interaction. A paired intervention shows that diversity modulation changes late-run replacement behaviour without a detectable effect on final error at the tested budget. The evidence supports LAO as a parsimonious evaluation-gating mechanism with regime-qualified competitiveness, rather than as a generally superior optimizer.

arXiv:2609.10588v1 Announce Type: new Abstract: This paper formalizes threshold-based selection as an evaluation-gating architecture in which each incumbent is tested before variation and a replacement is generated and evaluated only when contextual pressure exceeds intrinsic strength. The mechanism is instantiated as Leaf Abscission Optimization (LAO), using rank-based strength, a phenological seasonal signal, diversity modulation, environmental pressure, and a base regrowth kernel. A blocked two-to-the-fourth-power factorial analysis at dimension 10 on the CEC 2017 suite reduces the original multi-layer design to a parsimonious core: drift is harmful, while the other three auxiliary layers show no robust independent evidence of benefit. The resulting LAO-Core attains the third-best mean Friedman rank among nine optimizers at dimensions 10, 30, and 50 under the equal evaluation budget. A four-budget sweep shows budget-dependent relative performance, with adaptive differential-evolution baselines gaining relative advantage at larger budgets; the nine-cell dimension-budget analysis establishes neither an evaluation-budget-per-dimension-only law nor a statistically significant dimension-budget interaction. A paired intervention shows that diversity modulation changes late-run replacement behaviour without a detectable effect on final error at the tested budget. The evidence supports LAO as a parsimonious evaluation-gating mechanism with regime-qualified competitiveness, rather than as a generally superior optimizer.

2026-09-10T20:00:00+00:00