arXiv:2510.06492v2 Announce Type: replace Abstract: Latent world models are a promising approach for learning state representations and dynamics directly from high-dimensional observations, enabling robot control in hard-to-model settings. However, control performance ultimately depends on the latent representation encoding the required information for the task. In this work, we study latent-space safe control problems and show how partial observability can induce control failures when safety-relevant information is not preserved in the latent state. Specifically, we identify two world model failure modes: estimation gaps, where current observations do not reveal safety-critical quantities (e.g., temperature in a cooking task), and prediction gaps, where failures are observable once they occur but cannot be reliably anticipated from available observations. We introduce two diagnostics for these gaps: a mutual-information-based measure of safety observability and a rollout-based measure of future safety predictability. Finally, we present mitigation strategies for each failure mode: privileged multimodal supervision for estimation gaps and conformal risk calibration for prediction gaps. Across two hardware case studies -- using unimodal RGB world models and multimodal RGB+Tactile and RGB+Thermal variants -- we show that these mitigation strategies improve the safety of a Franka Research 3 manipulator on challenging cooking tasks under partial observability, albeit with increased conservativeness. More broadly, our work raises the question of when world model state representations are sufficient for reliable robot control
Science Journals
arXiv:2510.09783v2 Announce Type: replace Abstract: Oversampling is one of the most widely used approaches for addressing imbalanced classification. The core idea is to generate additional minority samples to rebalance the dataset. Most existing methods, such as SMOTE, require converting categorical variables into numerical vectors, which often leads to information loss. Recently, large language model (LLM)-based methods have been introduced to overcome this limitation. However, current LLM-based approaches typically generate minority samples with limited diversity, reducing robustness and generalizability in downstream classification tasks. To address this gap, we propose a novel LLM-based oversampling method designed to enhance diversity. First, we introduce a sampling strategy that conditions synthetic sample generation on both minority labels and features. Second, we develop a new permutation strategy for fine-tuning pre-trained LLMs. Third, we fine-tune the LLM not only on minority samples but also on interpolated samples to further enrich variability. Extensive experiments on 10 tabular datasets demonstrate that our method significantly outperforms eight SOTA baselines. The generated synthetic samples are both realistic and diverse. Moreover, we provide theoretical analysis through an entropy-based perspective, proving that our method encourages diversity in the generated samples.
arXiv:2510.10028v2 Announce Type: replace Abstract: The rapid advancement of Low-Altitude Economy Networks (LAENets) has enabled a variety of applications, including aerial surveillance, environmental sensing, and semantic data collection. To support these scenarios, unmanned aerial vehicles (UAVs) equipped with onboard vision-language models (VLMs) offer a promising solution for real-time multimodal inference. However, ensuring both inference accuracy and communication efficiency remains a significant challenge due to limited onboard resources and dynamic network conditions. In this paper, we first propose a UAV-enabled LAENet system model that jointly captures UAV mobility, user-UAV communication, and the onboard visual question answering (VQA) pipeline. Based on this model, we formulate a mixed-integer non-convex optimization problem to minimize task latency and power consumption under user-specific accuracy constraints. To solve the problem, we design a hierarchical optimization framework composed of two parts: (i) an Alternating Resolution and Power Optimization (ARPO) algorithm for resource allocation under accuracy constraints, and (ii) a Large Language Model-augmented Reinforcement Learning Approach (LLaRA) for adaptive UAV trajectory optimization. The large language model (LLM) serves as an expert in refining reward design of reinforcement learning in an offline fashion, introducing no additional latency in real-time decision-making. Numerical results demonstrate the efficacy of our proposed framework in improving inference performance and communication efficiency under dynamic LAENet conditions.
arXiv:2510.10448v2 Announce Type: replace Abstract: Search agents trained with reinforcement learning (RL) interleave reasoning with tool calls in a multi-turn, tool-integrated reasoning (TIR) loop, where each tool invocation returns an environment observation that is appended to the agent's context. As the rollout proceeds, these raw observations accumulate, inflating token cost and diluting the signal available for downstream reasoning. Unlike single-pass retrieve-then-read pipelines, where context compression is a one-time postprocessing step, the multi-turn RL setting requires compression that runs at every observation step while remaining decoupled from policy optimization. We introduce RECON (REasoning with CONdensation), a framework that addresses this challenge by inserting a dedicated observation compressor into the reasoning loop. The compressor is trained via a two-stage curriculum: relevance pretraining on QA datasets followed by multi-aspect distillation from proprietary LLMs, and remains frozen during RL training to preserve policy stability. Integrated into the Search-R1 search-agent pipeline, RECON reduces total context length by 35%, improves training speed by 5.4% and inference latency by 30.9%, while boosting average exact-match by 14.5% on the 3B agent and 3.0% on the 7B agent, with particular strength in multi-hop QA. These results establish learned observation compression as a key component for building practical, scalable RL-trained search agents.
arXiv:2510.12171v2 Announce Type: replace Abstract: Large Language Models have shown strong scientific reasoning ability, but their performance on materials science problems remains less studied. To fill this gap, we introduce MatSciBench, a comprehensive college-level benchmark comprising 1340 problems that span the essential subdisciplines of materials science. MatSciBench features a structured and fine-grained taxonomy that categorizes materials science questions into 6 primary fields and 31 subfields, together with a three-tier difficulty classification based on the reasoning length needed to solve each problem. MatSciBench includes detailed reference solutions for 946 questions, supports process-level error analysis, and contains 315 questions with images for evaluating multimodal reasoning. We evaluate leading thinking and non-thinking LLMs on MatSciBench, and further test three reasoning methods for non-thinking models: basic chain-of-thought prompting, tool augmentation, and self-correction. The results show that current models still face clear limits in college-level materials science reasoning. DeepSeek-R1 achieves the highest score on text-only questions at 75.22% accuracy, and GPT-5 performs the best on questions with images at 53.02%. Our analysis shows that tool augmentation improves many non-thinking models in a token-efficient way, while self-correction often fails to provide reliable gains and can revise correct answers into incorrect ones. We further analyze performance across difficulty levels, reasoning efficiency, multimodal reasoning, and failure patterns, and find that current models are mainly limited by domain knowledge gaps, calculation errors, problem comprehension failures, and difficulty in extracting precise information from scientific figures. Overall, MatSciBench provides a clear testbed for measuring current LLM limitations and guiding future work on scientific reasoning in materials science.
arXiv:2510.12705v3 Announce Type: replace Abstract: The reduction of a banded matrix to bidiagonal form is a critical step in the calculation of Singular Values, a cornerstone of scientific computing and AI. Although inherently parallel, this step has traditionally been considered unsuitable for GPUs due to its memory-bound nature. However, recent advances in GPU architectures, such as increased L1 memory per Streaming Multiprocessor or Compute Unit and larger L2 caches, have shifted this paradigm. In this work, we present the first GPU-accelerated algorithm for reducing a banded matrix to bidiagonal form, integrated into an open-source software package. Our algorithm builds on prior multicore CPU cache-efficient bulge-chasing methods, adapted to modern GPU architectures to optimize throughput. Leveraging Julia's high-level array abstractions and KernelAbstractions.jl, we implement a single function that is both hardware-agnostic and data-precision-aware, running efficiently across NVIDIA, AMD, Intel, and Apple Metal GPUs. We develop a hardware-aware performance model to guide tuning and identify key hyperparameters that govern optimal GPU performance for memory-bound workloads. We show that such workloads, when carefully optimized, can achieve substantial speed-ups on modern GPUs: our implementation outperforms multithreaded CPU libraries (PLASMA,SLATE) starting from matrix sizes as small as 1024x1024, and achieves over 100x speed-up on 32k x 32k matrices. Moreover, the algorithm's performance scales linearly with the matrix bandwidth, enabling efficient reduction of matrices with larger bandwidths, previously considered impractical.
arXiv:2510.13554v2 Announce Type: replace Abstract: The reasoning pattern of Large language models (LLMs) remains opaque, and reinforcement learning (RL) typically applies uniform credit across an entire generation, blurring the distinction between pivotal and routine steps. This work positions attention as a privileged substrate that renders the internal logic of LLMs legible, not merely as a byproduct of computation, but as a mechanistic blueprint of reasoning itself. We first distinguish attention heads between locally and globally focused information processing and reveal that locally focused heads produce a sawtooth pattern near the diagonal indicating phrasal chunks, while globally focused heads expose tokens that exert broad downstream influence over future tokens. We formalize these with two metrics: 1) Windowed Average Attention Distance, which measures the extent of backward attention within a clipped window; 2) Future Attention Influence, which quantifies a token's global importance as the average attention it receives from subsequent tokens. Taken together, these signals reveal a recurring preplan-and-anchor mechanism, where the model first performs a long-range contextual reference to generate an introductory token, which is immediately followed by or coincides with a semantic anchor token that organizes subsequent reasoning. Leveraging these insights, we introduce three novel RL strategies that dynamically perform targeted credit assignment to critical nodes (preplan tokens, anchor tokens, and their temporal coupling) and show consistent performance gains across various reasoning tasks. By aligning optimization with the model's intrinsic reasoning rhythm, we aim to transform opaque optimization into an actionable structure-aware process, hoping to offer a potential step toward more transparent and effective optimization of LLM reasoning.
arXiv:2510.20182v2 Announce Type: replace Abstract: Pedestrian simulation traditionally relies on expert-tuned, hand-crafted models that limit scalability and generalization. Meanwhile, large-scale video generation models have achieved high visual realism across diverse settings, motivating exploration of their potential as general-purpose world simulators. Existing benchmarks primarily assess single-subject realism rather than scenes with multiple interacting people, leaving the plausibility of multi-agent dynamics in generated videos untested. We propose a rigorous evaluation protocol to benchmark text-to-video (T2V) and image-to-video (I2V) models as implicit simulators of pedestrian dynamics. For I2V, we leverage start frames from established datasets to enable direct comparison with ground truth videos, while for T2V we design a prompt suite covering varied crowd densities and interaction types. A key component is a method to reconstruct 2D bird's-eye view trajectories from pixel-space without known camera parameters. Our analysis shows that leading models exhibit effective priors for plausible multi-agent behavior, though issues such as merging and disappearing pedestrians reveal limits to their physical consistency.
arXiv:2510.26253v3 Announce Type: replace Abstract: The ability to accurately interpret implied meanings plays a crucial role in human communication and language use, and language models are also expected to possess this capability. This study demonstrates that providing language models with pragmatic theories as prompts is an effective in-context learning approach for tasks to understand implied meanings. Specifically, we propose an approach in which an overview of pragmatic theories, such as Gricean pragmatics and Relevance Theory, is presented as a prompt to the language model, guiding it through a step-by-step reasoning process to derive a final interpretation. Experimental results showed that, compared to the baseline, which prompts intermediate reasoning without presenting pragmatic theories (0-shot Chain-of-Thought), our methods enabled language models to achieve up to 9.6\% higher scores on pragmatic reasoning tasks. Furthermore, we show that even without explaining the details of pragmatic theories, merely mentioning their names in the prompt leads to a certain performance improvement (around 1-3%) in larger models compared to the baseline.
arXiv:2511.04726v2 Announce Type: replace Abstract: To investigate the mechanical correlations between intraocular pressure (IOP) variations and glaucoma, this study presents a linear transversely isotropic poroelastic model of the lamina cribrosa (LC) based on Reissner-Mindlin plate theory. A key feature of the proposed framework is its analytical tractability, which allows the governing poroelastic equations to be solved in closed form under appropriate mechanical and hydraulic boundary conditions. Within this setting, linearity is used to capture the reversible component of the tissue response, providing a baseline description of the coupled solid-fluid feedback on which more complex time-dependent phenomena, such as viscoelastic effects and remodelling, may build.The results indicate that both strain and stress measures (in the form of shear strain and deviatoric stress measures) peak in the peripheral region of the LC, which is currently suspected to be the initial site of glaucomatous damage. These quantities increase with IOP, suggesting a pressure-dependent mechanical insult to the retinal ganglion cell (RGC) axons. In parallel, the model predicts a monotonic reduction in fluid content as IOP rises, which may contribute to ischemic phenomena and disc haemorrhages. The influence of material anisotropy was also examined, revealing that isotropic assumptions tend to overestimate the fluid content while underestimating shear strain. Given the current experimental challenges in measuring blood flow within the LC, the proposed model provides a valuable framework for exploring the coupled mechanical-hemodynamic behavior of the tissue and for inverse estimation of its mechanical parameters, such as the stiffness of the opening for the central retinal vessels.
arXiv:2512.00239v2 Announce Type: replace Abstract: The effectiveness of self-supervised learning (SSL) for physiological time series depends on the ability of a pretraining objective to preserve information about the underlying physiological state while filtering out unrelated noise. However, existing strategies are limited due to reliance on heuristic principles or poorly constrained generative tasks. To address this limitation, we propose a pretraining framework that exploits the information structure of a dynamical systems generative model across multiple time-series. This framework reveals our key insight that class identity can be efficiently captured by extracting information about the generative variables related to the system parameters shared across similar time series samples, while noise unique to individual samples should be discarded. Building on this insight, we propose PULSE, a cross-reconstruction-based pretraining objective for physiological time series datasets that explicitly extracts system information while discarding non-transferrable sample-specific ones. We establish theory that provides sufficient conditions for the system information to be recovered, and empirically validate it using a synthetic dynamical systems experiment. Furthermore, we apply our method to diverse real-world datasets, demonstrating that PULSE learns representations that can broadly distinguish semantic classes, increase label efficiency, and improve transfer learning.
arXiv:2512.03470v5 Announce Type: replace Abstract: A key challenge in infrared small target detection (IRSTD) is that weak target signal responses are easily obscured by strong background clutter, frequently resulting in missed detections. While traditional gradient-based methods attempt to capture fine details, their robustness is limited by the static fusion of multi-directional gradient features. In this paper, we rethink feature fusion from the perspective of Basis Decomposition Theory and propose a novel framework that reformulates the process into an explicit and adaptive decomposition-and-reconstruction paradigm. Specifically, we introduce the Basis Decomposition Module (BDM) and its specialized variant, the Gradient Decomposition Module (GDM) for IRSTD. GDMs treat the normalized gradient features as basis vectors to reconstruct a new feature, thereby maintaining detailed structures and highlighting infrared small targets. By integrating GDMs into a lightweight three-stage U-Net, we develop two unified architectures: the Spatial Gradient Basis Decomposition Network for single-frame detection and the Spatio-temporal Gradient Basis Decomposition Network for multi-frame scenarios. Extensive experiments demonstrate that our networks achieve state-of-the-art (SOTA) performance across multiple benchmarks, offering a superior balance between detection accuracy and computational efficiency. Our codes will be made public at: https://github.com/greekinRoma/IRSTD_HC_Platform.
arXiv:2512.07355v2 Announce Type: replace Abstract: Two traditions of interpretability have evolved side by side but seldom spoken to each other: Concept Bottleneck Models (CBMs), which prescribe what a concept should be, and Sparse Autoencoders (SAEs), which discover what concepts emerge. While CBMs use supervision to align activations with human-labeled concepts, SAEs rely on sparse coding to uncover emergent ones. We show that both paradigms instantiate the same geometric structure: each learns a set of linear directions in activation space whose nonnegative combinations form a concept cone. Supervised and unsupervised methods thus differ not in kind but in how they select this cone. Building on this view, we propose an operational bridge between the two paradigms. CBMs provide human-defined reference geometries, while SAEs can be evaluated by how well their learned cones approximate or contain those of CBMs. This containment framework yields quantitative metrics linking inductive biases -- such as SAE type, sparsity, or expansion ratio -- to emergence of plausible\footnote{We adopt the terminology of \citet{jacovi2020towards}, who distinguish between faithful explanations (accurately reflecting model computations) and plausible explanations (aligning with human intuition and domain knowledge). CBM concepts are plausible by construction -- selected or annotated by humans -- though not necessarily faithful to the true latent factors that organise the data manifold.} concepts. Using these metrics, we uncover a ``sweet spot'' in both sparsity and expansion factor that maximizes both geometric and semantic alignment with CBM concepts. Overall, our work unifies supervised and unsupervised concept discovery through a shared geometric framework, providing principled metrics to measure SAE progress and assess how well discovered concept align with plausible human concepts.
arXiv:2512.07998v2 Announce Type: replace Abstract: We present DIJIT, a novel binocular robotic head expressly designed for mobile agents that behave as active observers. DIJIT's unique breadth of functionality enables active vision research and the study of human-like eye and head-neck motions, their interrelationships, and how each contributes to visual ability. DIJIT is also being used to explore the differences between how human vision employs eye/head movements to solve visual tasks and current computer vision methods. DIJIT's design features nine mechanical degrees of freedom, while the cameras and lenses provide an additional four optical degrees of freedom. The ranges and speeds of the mechanical design are comparable to human performance. DIJIT attains 85\% of the peak human saccade speed. Our design includes the ranges of motion required for convergent stereo, namely, vergence, version, and cyclotorsion. Here, we present DIJIT and some aspects of its performance. We also present a novel method for saccadic camera movements, using a direct relationship between camera orientation and motor values. The resulting saccadic camera movements are close to human movements in terms of their accuracy, with 1.17$^\circ$ and 1.14$^\circ$ mean error for the left and right cameras, respectively.
arXiv:2512.08499v3 Announce Type: replace Abstract: Development of reliable and physically interpretable probabilistic frameworks for industrial prognostics remain nascent, and existing literature is often insensitive as inputs move away from the training manifold. In this paper, we develop two sampling-free, distance-aware physics-constrained probabilistic frameworks: (i) PC-SNGP and (ii) PC-SNER. Both apply spectral normalization to hidden layer weights, enforcing bi-Lipschitz distance-preserving representation from the input to the latent space. PC-SNGP replaces the dense output with Gaussian process whose posterior variance increases with input distance from the training manifold. PC-SNER modifies the output layer to predict Normal-Inverse-Gamma~(NIG) parameters for distance preserving estimation. To maintain balance between data fidelity and physical consistency during training, we introduce a dynamic weighting strategy for the physics-constrained loss. We also introduce a distance-aware-coefficient~(DAC) metric to quantify sensitivity to distributional shifts. Empirically, we validate both frameworks on rolling-element-bearings (REBs) prognostics using the PRONOSTIA, XJTU-SY, and HUST benchmark datasets. Experimental results demonstrate improved prediction accuracy and well-calibrated uncertainty estimates relative to competing baselines, while maintaining auditable performance in cross-validation and robustness under extreme adversarial perturbations.
arXiv:2512.12320v2 Announce Type: replace Abstract: Conventional soft pneumatic actuators, typically based on hollow elastomeric chambers, often suffer from small structural support and require costly geometry-specific redesigns for multimodal functionality. Porous materials such as foam, filled into chambers, can provide structural stability for the actuators. However, methods to achieve programmable deformation by tailoring the porous body itself remain underexplored. In this paper, a novel design method is presented to realize soft porous actuators with programmable deformation by incising specific patterns into the porous foam body. This approach introduces localized structural anisotropy of the foam guiding the material's deformation under a global vacuum input. Furthermore, three fundamental patterns on a cylindrical foam substrate are discussed: transverse for bending, longitudinal for tilting, and diagonal for twisting. A computational model is built with Finite Element Analysis (FEA), to investigate the mechanism of the incision-patterning method. Experiments demonstrate that with a potential optimal design of the pattern array number N, actuators can achieve bending up to $80^{\circ}$ (N=2), tilting of $18^{\circ}$ (N=1), and twisting of $115^{\circ}$ (N=8). The versatility of our approach is demonstrated via pattern transferability, scalability, and mold-less rapid prototyping of complex designs. As a comprehensive application, we translate the human hand crease map into a functional incision pattern, creating a bio-inspired soft robot hand capable of human-like adaptive grasping. Our work provides a new, efficient, and scalable paradigm for the design of multi-functional soft porous robots.
arXiv:2601.06077v2 Announce Type: replace Abstract: This work aims to rigorously define the values of perception, prediction, communication, and common sense in decision making. The defined quantities are decision-theoretic, but have information-theoretic analogues, e.g., they share some simple but key mathematical properties with Shannon entropy and mutual information, and can reduce to these quantities in particular settings. One interesting observation is that, the value of perception without prediction can be negative, while the value of perception together with prediction and the value of prediction alone are always nonnegative. The defined quantities suggest answers to practical questions arising in the design of autonomous decision-making systems. Example questions include: Do we need to observe and predict the behavior of a particular agent? How important is it? What is the best order to observe and predict the agents? The defined quantities may also provide insights to cognitive science and neural science, toward the understanding of how natural decision makers make use of information gained from different sources and operations.
arXiv:2601.06188v3 Announce Type: replace Abstract: As Earth-observing satellite constellations grow in size and capability, distributed onboard control offers a pathway to novel responses and time-sensitive measurements. However, deploying autonomy to satellites requires efficient computation and communication. This work addresses the challenge of scheduling observations for hundreds of satellites in a dynamic, large-scale problem with millions of variables. We present the dynamic multi-satellite constellation observation scheduling problem (DCOSP), a new formulation of dynamic distributed constraint optimization problems (DDCOP) that models integrated scheduling and execution. DCOSP features a novel optimality condition, for which we construct an exact omniscient offline algorithm. Motivated by the strong resource constraints of onboard satellite operations, we introduce a framework to incorporate metareasoning in DDCOPs that controls when agents expend resources to recompute solutions. In addition, we present the dynamic incremental neighborhood stochastic search (D-NSS) algorithm, an incomplete online decomposition-based DDCOP algorithm that repairs localized sub-problems in response to dynamic events. We demonstrate in realistic simulations that D-NSS converges to near-optimal solutions, outperforming standard DDCOP baselines in solution quality, computation time, and message volume, while our metareasoning framework successfully balances resource conservation with utility. As part of the NASA FAME mission, this work lays the foundation for the largest in-space demonstration of distributed multi-agent AI to date.
arXiv:2601.12263v2 Announce Type: replace Abstract: Vision-Language Models (VLMs) integrate visual and textual knowledge into unified representations that increasingly underpin modern retrieval and recommendation systems. However, it remains unclear how reliably these models utilize their cross-modal knowledge when ranking multimodal items, and whether their knowledge grounding can be subverted. In this paper, we expose a fundamental vulnerability in how VLMs apply multimodal knowledge for product ranking: through Multimodal Generative Engine Optimization (MGEO), we show that an adversary can manipulate a VLM's ranking decisions by jointly crafting imperceptible image perturbations and fluent textual suffixes that exploit the model's internal cross-modal knowledge coupling. Using an alternating optimization strategy, MGEO targets the deep interactions between visual and linguistic representations within the VLM, achieving rank manipulations that substantially exceed those of unimodal attacks and heuristic baselines powered by strong commercial models. Our findings reveal that surface-level content quality is insufficient for rank promotion; instead, direct alignment with the model's internal knowledge utilization mechanism is required. These results raise important questions on the faithfulness and robustness of knowledge grounding in multimodal foundation models, and motivate future work on defense mechanisms for multimodal retrieval systems. Code is available at: https://github.com/glad-lab/MGEO
arXiv:2601.15408v2 Announce Type: replace Abstract: Medical vision-language models can automate the generation of radiology reports but struggle with accurate visual grounding and factual consistency. Existing models often misalign textual findings with visual evidence, leading to unreliable or weakly grounded predictions. We present CURE, an error-aware curriculum learning framework that improves grounding and report quality without any additional data. CURE fine-tunes a multimodal instructional model on phrase grounding, grounded report generation, and anatomy-grounded report generation using public datasets. The method dynamically adjusts sampling based on model performance, emphasizing harder samples to improve spatial and textual alignment. CURE improves grounding accuracy by +0.35 IoU, boosts report quality by +0.192 CXRFEScore, and reduces hallucinations by 18.6%. CURE is a data-efficient framework that enhances both grounding accuracy and report reliability. Code is available at https://github.com/PabloMessina/CURE and model weights at https://huggingface.co/pamessina/medgemma-4b-it-cure
arXiv:2601.15727v3 Announce Type: replace Abstract: The performance of modern AI systems is fundamentally constrained by the quality of their underlying GPU kernels, which translate high-level algorithmic semantics into low-level hardware operations. Achieving near-optimal kernels requires expert-level understanding of hardware architectures and programming models, making kernel engineering a critical but notoriously time-consuming and non-scalable process. Recent advances in large language models and LLM-based agents have opened new possibilities for automating kernel generation and optimization. LLMs are well-suited to compress expert-level kernel knowledge that is difficult to formalize, while agentic systems further enable scalable optimization by casting kernel development as an iterative, feedback-driven loop. Rapid progress has been made in this area. However, the field remains fragmented and lacks a systematic perspective for LLM-driven kernel generation. This survey addresses this gap by providing a structured overview of existing approaches, spanning LLM-based approaches and agentic optimization workflows, and systematically organizing the datasets and benchmarks that underpin learning and evaluation in this domain. Moreover, key open challenges and future research directions are further outlined, aiming to establish a comprehensive reference for the next generation of automated kernel optimization. To keep track of this field, we maintain an open-source GitHub repository at https://github.com/flagos-ai/awesome-LLM-driven-kernel-generation.
arXiv:2601.17957v4 Announce Type: replace Abstract: Grassroots Logic Programs (GLP) is a concurrent logic programming language in which logic variables are partitioned into paired readers and writers. An assignment is produced at most once via a writer and consumed at most once via its paired reader, and may contain additional readers and/or writers. This enables the concise expression of rich multidirectional communication modalities. ``Logic Programs as Types for Logic Programs'' (LICS'91) defined types as regular sets of paths over the Herbrand atom semantics of a logic program. Here, we develop a \emph{moded-atom semantics} that extends the standard Herbrand atom semantics in two ways: (\ia)~each atom subterm carries a \emph{mode}, recording whether it is consumed from or produced to the environment; and (\ib)~partial computations, including those that deadlock, fail, or never terminate, also contribute moded atoms to the semantics. We define types to be regular sets of \emph{moded paths} over this semantics, give a syntactic definition of GLP well-typing, and prove that a well-typed program is sound: every output path in its well-typed moded-atom semantics conforms to its declared output type. A type checker for GLP was implemented \emph{by} AI (Claude) in Dart, starting from the mathematical specification of Typed GLP (this paper), deriving from it an English+pseudocode spec (written by AI), and from the spec deriving Dart code (by AI). While GLP is naturally untyped, the motivation for typing it was \emph{for} AI: tasking AI to program complex communication modalities and hoping for the best turned out to be a tenuous strategy. The discipline we developed with Typed GLP is for the human designer and AI to jointly develop formal GLP type definitions and declarations, together with informal intent of the declared procedures, and only then let AI write the GLP code.
arXiv:2601.21522v2 Announce Type: replace Abstract: The performance of large language models (LLMs) on verifiable tasks is usually measured by pass@k, the probability of answering a question correctly at least once in k trials. At a fixed budget, a more suitable metric is coverage@cost, the average number of unique questions answered as a function of the total number of attempts. We connect the two metrics and show that the empirically-observed power-law behavior in pass@k leads to a sublinear growth of the coverage@cost (diminishing returns). To solve this problem, we propose Reset-and-Discard (ReD), a query method of LLMs that increases coverage@cost for a given budget, regardless of the pass@k form. Moreover, given a pass@k, we can quantitatively predict the savings in the total number of attempts using ReD. If pass@k is not available for the model, ReD can infer its power-law exponent. Experiments on three LLMs across coding (HumanEval), math (GSM8K), and reasoning (MMLU-Pro) benchmarks demonstrate that ReD substantially reduces the required attempts, tokens, and USD cost to reach a desired coverage, while also offering an efficient way to measure inference power-laws. ReD's advantage is maintained for imperfect verifiers and outperforms the tested allocation baselines.
arXiv:2601.21816v2 Announce Type: replace Abstract: Evaluating the performance of large language models (LLMs) from human preference data is crucial for obtaining LLM leaderboards. However, many existing approaches either rely on restrictive parametric assumptions or lack valid uncertainty quantification when flexible machine learning methods are used. In this paper, we propose a nonparametric statistical framework, called DMLRank, for comparing and ranking LLMs from preference data using debiased machine learning (DML). For this, we introduce generalized average ranking scores (GARS), which generalize commonly used ranking models, including the Bradley-Terry model or PageRank/ Rank centrality, with complex human responses such as ties. DMLRank comes with the following advantages: (i)~It produces statistically efficient estimates of GARS ranking scores. (ii) It naturally allows the incorporation of black-box machine learning methods for estimation. (iii) It can be combined with pre-trained LLM evaluators (e.g., using LLM-as-a-judge). (iv) It suggests optimal policies for collecting preference data under budget constraints. We demonstrate these advantages both theoretically and empirically using both synthetic and real-world preference datasets. In summary, our framework provides practitioners with powerful, state-of-the-art methods for comparing or ranking LLMs for leaderboards.
arXiv:2602.10016v3 Announce Type: replace Abstract: Deriving predictable scaling laws that govern the relationship between model performance and computational investment is crucial for designing and allocating resources in massive-scale recommendation systems. While such laws are established for large language models, they remain challenging for recommendation systems, especially those processing both user history and context features. We identify poor scaling efficiency as the main barrier to predictable power-law scaling, stemming from inefficient modules with low Model FLOPs Utilization (MFU) and suboptimal resource allocation. We introduce Kunlun, a scalable architecture that systematically improves model efficiency and resource allocation. Our low-level optimizations include Generalized Dot-Product Attention (GDPA), Hierarchical Seed Pooling (HSP), and Sliding Window Attention. Our high-level innovations feature Computation Skip (CompSkip) and Event-level Personalization. These advances increase MFU from 17% to 37% on NVIDIA B200 GPUs and double scaling efficiency over state-of-the-art methods. Kunlun is now deployed in major Meta Ads models, delivering significant production impact.