Topic
optimization
Optimal Scheduling in QA Forums Could Boost Knowledge Worker Efficiency, New Research Shows
Researchers Negi, Rohit, Yilmaz, and Mustafa present a model for optimal scheduling in question-answering forums where knowledge workers answer requests. They calculate system capacity and design schedulers to keep the system stable, also exploring how collaboration can increase capacity.
Multi-Agent RL System MAMO Automates Weight Selection for Constrained Optimization Problems
MAMO decouples task execution from objective design using multi-agent RL to automatically select reward weights for constrained optimization, improving adaptability in dynamic environments.
Logistics Flatbed Freight Poised for Surge as AI Data Centers and Military Spending Boost Demand
According to FreightWaves, the flatbed freight market may have seen a recent cool-off, but massive investments in AI data centers and a record-breaking defense bill are set to fuel unprecedented demand. Flatbed trucking will be essential for moving concrete for new facilities and munitions for global defense, putting capacity and rates under pressure.
Residual-Space Evolutionary Optimization via Flow-based Generative Models
A new framework called residual-space evolutionary optimization addresses the challenge of data editing with non-differentiable objectives in flow-based generative models. By operating in residual space, it separates local exploitation (self-pollination) from broader exploration (cross-pollination). The method was validated on the MorphoMNIST benchmark and crystal data, showing balanced target alignment, instance preservation, and diversity.
Learning What to Remember: Observability-Safe Memory Retention via Constrained Optimization for Long-Horizon Language Agents
A new research paper formulates memory retention in long-horizon language agents as a constrained stochastic optimization problem, proposing OSL-MR (Observability-Safe Learning for Memory Retention). The method combines an evidence learner with a Mixed-Score heuristic, achieving superior performance under tight budgets on benchmarks LoCoMo and LongMemEval. The work establishes a principled foundation for memory management in AI agents.
AIGB-Pearl: New AI Method Combines Generative Planning and Policy Optimization for Auto-bidding
Researchers propose AIGB-Pearl, a method that integrates generative planning and policy optimization to overcome performance bottlenecks in AI-generated auto-bidding. The approach uses a trajectory evaluator and a KL-Lipschitz-constrained score-maximization scheme for safe exploration beyond offline datasets, demonstrating state-of-the-art results in simulated and real-world advertising systems.
AutoPass: Evidence-Guided LLM Agents Achieve Compiler Speedups of 1.117x on ARM64
Researchers present AutoPass, a multi-agent LLM framework that uses compiler and runtime evidence to guide compiler optimization decisions. Without training, it outperforms expert-tuned heuristics and classical autotuning, achieving geometric-mean speedups of 1.043x on x86-64 and 1.117x on ARM64 over LLVM -O3.
New Algorithm SLARouter Cuts LLM Inference Costs by 2.2x While Guaranteeing SLA Compliance
A new research paper introduces SLARouter, an online routing algorithm that learns a cost-optimal policy from sparse user feedback while guaranteeing SLA compliance. Experiments show up to 2.2x cost reduction over existing baselines without per-benchmark tuning.
Zero-Inflated Gaussian EDAs Enable Parameter-Space Sparsity Without Hand-Crafted Operators
Researchers propose multivariate zero-inflated Gaussian distributions as sampling laws for estimation-of-distribution algorithms (EDAs), enabling parameter-space sparsity without hand-crafted operators. The ZIG-EDA jointly optimizes sparsity patterns and active values, and on the Lunar Lander benchmark it converges faster and achieves higher returns than dense and sparse alternatives.
RACL: Reasoning-Agent Control Layers Show Promise for Continuous Metaheuristic Learning in Vehicle Routing
A new research paper introduces Reasoning-Agent Control Layers (RACL), a method that places an AI reasoning agent above existing metaheuristic optimizers to discover and validate control rules. Tested on vehicle routing, RACL outperformed baseline policies in 21 of 21 cases, achieving average cost reductions of up to 8.3%.
FastMix: Gradient-Based Data Mixture Optimization Reduces Search Cost in AI Training
FastMix is a novel framework that automates data mixture discovery by training only a single proxy model and jointly optimizing mixture coefficients and model parameters via gradient descent. It reformulates mixture selection as a bilevel optimization problem, enabling efficient, scalable optimization that outperforms baselines.
New AI Training Method Reduces Decision Errors in Stochastic Optimization for Supply Chain and Finance
Researchers propose Decision-Weighted Flow Matching (DW-FM), a training framework for conditional generative models that minimizes decision regret rather than distributional error. The method improves performance on contextual stochastic optimization tasks including portfolio optimization, financial planning, and traffic CVaR, which have direct applications in supply chain and logistics under uncertainty.
OBCache Prunes KV Cache for Efficient Long-Context LLM Inference with Output-Aware Scoring
A new method called Optimal Brain Cache (OBCache) treats key-value cache eviction as a layer-wise structured pruning problem. By measuring token saliency through perturbation in attention outputs, OBCache outperforms heuristic-based approaches on LLaMA and Qwen models, consistently improving long-context accuracy according to the paper.
Pruning Optimisations Boost LUT-Based Neural Network Scalability and Efficiency
Researchers propose a pruning-optimised Look-Up Table (LUT) matrix multiplication unit (LUT-MU) to address scalability limits in LUT-based neural networks. Deployed on FPGAs, it delivers up to 1.6x throughput improvement and 4.2x energy efficiency gains over CUDA-based implementations, with 1.3 to 2.6x resource savings versus original MADDNESS-based networks.
PASTE System Cuts AI Agent Latency by 43.5% via Parallel Tool Execution and LLM Generation
A new system called PASTE reduces average task completion time for AI agents by 43.5% by parallelizing tool execution with LLM generation. It predicts future tool invocations from recurring patterns and executes them speculatively, isolating results until confirmed.
AgenticRec: A Recommender Framework That Aligns LLM Reasoning with User Preferences
Researchers propose AgenticRec, a framework that treats recommendation as a tool-integrated reasoning process. It employs a two-stage training paradigm to overcome misalignment between LLM reasoning trajectories and recommendation feedback, improving fine-grained preference distinction.
New Book on Optimal Transport Offers Machine Learning Practitioners a Unified Framework
A new book titled 'Optimal Transport for Machine Learners' presents a comprehensive overview of optimal transport techniques tailored for machine learning. It covers key concepts such as Kantorovich couplings, Wasserstein distances, Sinkhorn scaling, and gradient flows, providing a mathematical framework for comparing probability measures in ML applications.
SDS-LoRA: New Low-Rank Adaptation Method Fixes Gradient Distortion in Large Model Fine-Tuning
A new paper on arXiv introduces SDS-LoRA, a low-rank parameterization that overcomes anisotropic gradient scaling in LoRA. By structurally decoupling singular values from the backward pass, SDS-LoRA ensures gradients are only applied through orthonormal bases, improving convergence and reducing the performance gap to full fine-tuning. Experimental results across natural language and vision benchmarks show enhanced adaptation performance.
New Fluid-Guided Algorithm Optimizes LLM Inference Scheduling Under Memory Constraints
A new paper from researchers including David Simchi-Levi introduces a fluid-guided online scheduling approach for LLM inference that addresses memory constraints from Key-Value cache growth. The WAIT and Nested WAIT algorithms approximate an optimal fluid benchmark, reducing latency in overloaded regimes according to simulations on Llama-2-7B with A100 GPUs.
New Architecture GRIL Enables Gradient Descent-Like Learning in Linear Recurrent Networks
Researchers introduce the Gradient-based Recurrent In-context Learner (GRIL), a linear recurrent network architecture with windowed cross-product self-attention that can implement minibatch gradient descent on a task-specific predictor in a single forward pass. The design achieves strong performance on synthetic in-context learning tasks, Long Range Arena, and language modeling.
HOLO-MPPI Framework Promises Robust Motion Planning for Autonomous Robots Without Per-Scenario Tuning
HOLO-MPPI is a new motion planning framework that combines hierarchical policy learning with stochastic optimal control. It addresses the brittleness of end-to-end reinforcement learning and the scalability issues of manually designed priors for MPPI. Tested in autonomous driving scenarios, it outperforms baselines while maintaining real-time control.
SMEPilot Boosts LLM Inference Up to 3.94x on CPUs with Scalable Matrix Extensions
Researchers have developed SMEPilot, an LLM inference engine that leverages Arm Scalable Matrix Extension (SME) to optimize execution on CPUs. By selecting CPU-only, SME-only, or cooperative SME+CPU execution per operator shape, SMEPilot improves end-to-end inference by up to 3.94x across multiple models and platforms.
DOG-DPO: Training-Free Geometric Data Selection Boosts LLM Safety Alignment with 11% of Data
Researchers propose DOG-DPO, a training-free data selection framework for LLM safety alignment that treats preference pairs as geometric directions. By decomposing multi-dataset geometry and maximizing diversity-based coverage, it achieves strong utility-robustness trade-off using only 11% of preference pairs, recovering most safety gains of full-data training while being teacher-free, training-free, and substantially faster than traditional selection methods.
Embedded Arena: How Hardware Feedback Lets an LLM Agent Automate Edge AI Optimization
Researchers introduce Embedded Arena, a hardware-in-the-loop system where an LLM agent iteratively refines model and firmware by compiling, flashing, and measuring on real microcontrollers. Without hardware feedback, frontier models like Claude Opus 4.7 and Gemini 3.1 Pro fail entirely, but the closed-loop approach achieves first successful deployment in three iterations and surpasses human experts in seven. The method yields 250x compression for vision models and 400x for audio, enabling battery-free operation on commercial MCUs.
LearnOpt Uses Knowledge Graphs and Optimization to Reveal Hidden Structure in Standardized Exams
Researchers introduce LearnOpt, a system that recovers latent cognitive structures from standardized examinations using knowledge graphs and constrained optimization. Applied to NEET and JEE Advanced, it reveals stable skill distributions within syllabus regimes and significant shifts after curricular changes.
StarOR: New AI Framework Combines Tree Search and Reinforcement Learning for Optimization Modeling
A new AI framework called StarOR combines Monte Carlo Tree Search with test-time reinforcement learning to solve hierarchical optimization modeling problems. It decomposes modeling into four stages, uses a LoRA adapter updated via GRPO, and achieves state-of-the-art results on five benchmarks with a 4B parameter backbone, outperforming existing methods and frontier LLMs.
Proximal Policy Optimization Achieves Faster Convergence in Discrete Sampling Research
A new paper on arXiv explores policy gradient algorithms for training stochastic policies under the Generative Flow Network (GFlowNet) framework. The authors derive equivalents of standard policy gradient algorithms and, for the first time, successfully apply proximal policy optimization (PPO) to GFlowNets, demonstrating improved convergence speed and data efficiency on benchmarks including synthetic energies and molecular graph generation.
Spokes Optimizes Diverse Pretraining Data Selection for LLMs, Boosting Performance
Researchers introduce Spokes, a method that directly optimizes diversity in pretraining data selection for large language models. Using a probabilistic framework based on the G-Vendi score and exponentiated gradient descent, Spokes achieves significantly more diverse subsets and improves downstream performance by up to 1.5 points over random sampling.
Structural Heterogeneity in LLM Verification: Signal Quality Varies Across Cost Strata
A recent paper on arXiv identifies a fundamental failure mode in LLM verification: uncertainty signals are heteroskedastic across cost strata, with some error-concentrating regions exhibiting near-random discriminability. The authors propose a cost-stratified thresholding intervention (CST) that improves hit rate by up to 17 percentage points without gradient updates, showing that structural heterogeneity, not optimizer weakness, is the primary bottleneck.
New Hardware-Aware Neural Architecture Search Runs on Embedded Devices with Under 512MB RAM
Researchers propose a hardware-aware neural architecture search (HW NAS) method that runs on embedded devices with under 512MB of RAM. It produces tiny convolutional neural networks for low-end microcontrollers, enabling on-device AI without cloud dependence. The approach achieves state-of-the-art results on the Visual Wake Word dataset.
A Framework for Governing Optimization in AI Systems: Architectural Wisdom
The paper 'Architectural Wisdom' argues that modern AI failures stem from optimizing underspecified objectives, not lack of intelligence. It proposes a corrigible objective-governance layer above the optimization substrate, made of four components and a six-coordinate wisdom tuple. The framework is motivated by eight cases of contemporary AI failures and aims to prevent harmful outcomes.
E-mem: Multi-Agent Framework for Episodic Memory Reconstruction Boosts LLM Reasoning Efficiency by 70%
Researchers propose E-mem, a multi-agent framework that reconstructs episodic context for LLM agent memory, inspired by biological engrams. It uses a hierarchical architecture with assistant agents maintaining uncompressed contexts and a master agent orchestrating planning, achieving 54% F1 on the LoCoMo benchmark, surpassing the state-of-the-art GAM by 7.75% with over 70% token cost reduction.
daVinci-kernel: Reinforcement Learning Framework Automates GPU Kernel Optimization with Co-Evolving Skill Library
A new reinforcement learning framework called daVinci-kernel automates GPU kernel optimization by co-evolving skill selection, summarization, and utilization. The framework, detailed in a preprint on arXiv, uses three agents sharing one LLM backbone and achieves 37.2%, 70.6%, and 32.2% on KernelBench Level 1, 2, and 3 respectively, outperforming prior RL-trained models.
Technology How Apple's iOS 27 Boosts Older iPhone Speed and Extends Device Longevity
Apple's upcoming iOS 27 update brings platform stability improvements that make older iPhones, including the iPhone 11 from 2019, run faster and more responsively. Key gains include up to 30% faster app launches, 80% faster AirDrop transfers, and 70% faster photo library loading, achieved through CPU scheduler optimizations. The move supports Apple's services business—which hit $31 billion in revenue—and reinforces brand loyalty, though older devices miss new Apple Intelligence features.
Technology How Linear Achieves Millisecond Response Times: A Technical Breakdown for Enterprise Decision-Makers
Linear's web app updates issues in milliseconds by treating IndexedDB as the primary database, applying mutations locally before syncing via WebSocket. Co-founder Tuomas built the sync engine from day one. For CTOs evaluating performance, this approach eliminates network bottlenecks and loading states.