Agentic reinforcement learning has advanced large language models (LLMs) to reason through long chain-of-thought trajectories while interleaving external tool use. Existing approaches assume a fixed inventory of tools, which limits the adaptability of LLM agents to new or evolving toolsets. We present AutoTool, a training framework that equips LLM agents with dynamic tool-selection capabilities throughout their reasoning trajectories. AutoTool employs a dual-phase optimization pipeline: (i) SFT and RL-based trajectory stabilization for coherent reasoning, and (ii) KL-regularized Plackett-Luce Ranking to refine consistent multi-step tool selection. We further build a 200k dataset with explicit tool-selection rationales across 1,000+ tools and 100+ tasks spanning mathematics, science, code generation, and multimodal reasoning. Across ten diverse benchmarks, we train two base models, Qwen3-8B and Qwen2.5-VL-7B, with AutoTool. With fewer parameters, AutoTool consistently outperforms advanced LLM agents and tool-integration methods, yielding average gains of 6.4% in math & science reasoning, 4.5% in search-based QA, 7.7% in code generation, and 6.9% in multimodal understanding. In addition, AutoTool exhibits stronger generalization by dynamically leveraging unseen tools from evolving toolsets during inference.

CLIP delivers strong zero-shot classification but remains highly vulnerable to adversarial attacks. Prior adversarial fine-tuning work primarily matches predicted logits between clean and adversarial examples, which overlooks uncertainty calibration and may degrade the zero-shot generalization. A common expectation in reliable uncertainty estimation is that predictive uncertainty should increase as inputs become more difficult or shift away from the training distribution. However, we frequently observe the opposite in the adversarial setting: perturbations not only degrade accuracy but also suppress uncertainty, leading to severe miscalibration and over-confidence. This reveals a critical reliability gap beyond robustness. To bridge this gap, we propose an adversarial fine-tuning objective for CLIP considering both accuracy and uncertainty. By reparameterizing CLIP outputs as the concentration parameters of a Dirichlet distribution, we propose a unified representation that captures relative semantic structure and confidence magnitude. This enables holistic distribution alignment under perturbations, moving beyond single-logit anchoring and restoring calibrated uncertainty. Experiments across multiple zero-shot benchmarks demonstrate that our method significantly improves uncertainty calibration and achieves competitive adversarial robustness while preserving clean accuracy.

Few-shot semantic segmentation (FSS) aims to segment novel classes in query images using only a small annotated support set. While prior research has mainly focused on improving decoders, the encoder's limited ability to extract meaningful features for unseen classes remains a key bottleneck. In this work, we introduce \textit{Take a Peek} (TaP), a simple yet effective method that enhances encoder adaptability for both FSS and cross-domain FSS \rev{by inducing a lightweight \textit{feature-space shift} conditioned on the support set}. TaP leverages Low-Rank Adaptation to fine-tune the encoder on the support set with minimal computational overhead, enabling fast adaptation to novel classes while mitigating catastrophic forgetting. Our method is model-agnostic and can be seamlessly integrated into existing FSS pipelines. Extensive experiments across multiple benchmarks--including COCO $20^i$, Pascal $5^i$, and cross-domain datasets such as DeepGlobe, ISIC, and Chest X-ray--demonstrate that TaP consistently improves segmentation performance across diverse models and shot settings. Notably, TaP delivers significant gains in complex multi-class scenarios, highlighting its practical effectiveness in realistic settings. A rank sensitivity analysis also shows that strong performance can be achieved even with low-rank adaptations, thereby ensuring computational efficiency. By addressing a critical limitation in FSS--the encoder's generalization to novel classes--TaP paves the way toward more robust, efficient, and generalizable segmentation systems. The code is available at https://github.com/pasqualedem/TakeAPeek.

This article presents the creation of an Estonian-language dataset for document-level subjectivity, analyzes the resulting annotations, and reports an initial experiment of automatic subjectivity analysis using a large language model (LLM). The dataset comprises of 1,000 documents-300 journalistic articles and 700 randomly selected web texts-each rated for subjectivity on a continuous scale from 0 (fully objective) to 100 (fully subjective) by four annotators. As the inter-annotator correlations were moderate, with some texts receiving scores at the opposite ends of the scale, a subset of texts with the most divergent scores was re-annotated, with the inter-annotator correlation improving. In addition to human annotations, the dataset includes scores generated by GPT-5 as an experiment on annotation automation. These scores were similar to human annotators, however several differences emerged, suggesting that while LLM based automatic subjectivity scoring is feasible, it is not an interchangeable alternative to human annotation, and its suitability depends on the intended application.

This paper presents AIDEN, an artificial intelligence-based assistant designed to enhance the autonomy and daily quality of life of visually impaired individuals, who often struggle with object identification, text reading, and navigation in unfamiliar environments. Existing solutions such as screen readers or audio-based assistants facilitate access to information but frequently lead to auditory overload and raise privacy concerns in open environments. AIDEN addresses these limitations with a hybrid architecture that integrates You Only Look Once (YOLO) for real-time object detection and a Large Language and Vision Assistant (LLaVA) for scene description and Optical Character Recognition (OCR). A key novelty of the system is a continuous haptic guidance mechanism based on a Geiger-counter metaphor, which supports object centering without occupying the auditory channel, while privacy is preserved by ensuring that no personal data are stored. Empirical evaluations with visually impaired participants assessed perceived ease of use and acceptance using the Technology Acceptance Model (TAM). Results indicate high user satisfaction, particularly regarding intuitiveness and perceived autonomy. Moreover, the ``Find an Object'' achieved effective real-time performance. These findings provide promising evidence that multimodal haptic-visual feedback can improve daily usability and independence compared to traditional audio-centric methods, motivating larger-scale clinical validations.

Neural prediction offers a promising approach to forecasting the individual variability of neurocognitive functions and disorders and providing prognostic indicators for personalized invention. However, it is challenging to translate neural predictive models into medical artificial intelligent applications due to the limitations of domain shift and label scarcity. Here, we propose the directed evolution model (DEM), a novel computational model that mimics the trial-and-error processes of biological directed evolution to approximate optimal solutions for predictive modeling tasks. We demonstrated that the directed evolution algorithm is an effective strategy for uncertainty exploration, enhancing generalization in reinforcement learning. Furthermore, by incorporating replay buffer and continual backpropagate methods into DEM, we provide evidence of achieving better trade-off between exploitation and exploration in continuous learning settings. We conducted experiments on four different datasets for children with cochlear implants whose spoken language developmental outcomes vary considerably on the individual-child level. Preoperative neural MRI data has shown to accurately predict the post-operative outcome of these children within but not across datasets. Our results show that DEM can efficiently improve the performance of cross-domain pre-implantation neural predictions while addressing the challenge of label scarcity in target domain.

Interchanges are crucial nodes for vehicle transfers between highways, yet the lack of real-time ramp detectors creates blind spots in traffic prediction. To address this, we propose a Spatio-Temporal Decoupled Autoencoder (STDAE), a two-stage framework that leverages cross-modal reconstruction pretraining. In the first stage, STDAE reconstructs historical ramp flows from mainline data, forcing the model to capture intrinsic spatio-temporal relations. Its decoupled architecture with parallel spatial and temporal autoencoders efficiently extracts heterogeneous features. In the prediction stage, the learned representations are integrated with models such as GWNet to enhance accuracy. Experiments on three real-world interchange datasets show that STDAE-GWNET consistently outperforms thirteen state-of-the-art baselines and achieves performance comparable to models using historical ramp data. This demonstrates its effectiveness in overcoming detector scarcity and its plug-and-play potential for diverse forecasting pipelines.

Conformal Prediction (CP) has emerged as a powerful statistical framework for high-stakes classification applications. Instead of predicting a single class, CP generates a prediction set, guaranteed to include the true label with a pre-specified probability. The performance of different CP methods is typically assessed by their average prediction set size. In setups where the classes can be partitioned into semantic groups, e.g., diseases that require similar treatment, users can benefit from prediction sets that are not only small on average, but also contain a small number of semantically different groups. This paper begins by addressing this problem and ultimately offers a widely applicable tool for boosting any CP method on any dataset. First, given a class partition, we propose augmenting the CP score function with a term that penalizes predictions with out-of-group errors. We theoretically analyze this strategy and prove its advantages for group-related metrics. Surprisingly, we show mathematically that, for common class partitions, it can also reduce the average set size of any CP score function. Our analysis reveals the class-similarity factors behind this improvement and motivates a variant that can further reduce prediction set size by leveraging the model's embeddings, without requiring any human semantic partition. Finally, we present an extensive empirical study, encompassing prominent CP methods, multiple models, and several datasets, which demonstrates that our class-similarity-based approach consistently enhances CP methods.

Grasping in a densely cluttered environment is a challenging task for robots. Previous methods tried to solve this problem by actively gathering multiple views before grasp pose generation. However, they either overlooked the importance of the grasp distribution for information gain estimation or relied on the projection of the grasp distribution, which ignores the structure of grasp poses on the SE(3) manifold. To tackle these challenges, we propose a calibrated energy-based model for grasp pose generation and an active view selection method that estimates information gain from grasp distribution. Our energy-based model captures the multi-modality nature of grasp distribution on the SE(3) manifold. The energy level is calibrated to the success rate of grasps so that the predicted distribution aligns with the real distribution. The next best view is selected by estimating the information gain for grasp from the calibrated distribution conditioned on the reconstructed environment, which could efficiently drive the robot to explore affordable parts of the target object. Experiments on simulated environments and real robot setups demonstrate that our model could successfully grasp objects in a cluttered environment with limited view budgets compared to previous state-of-the-art models. Our simulated environment can serve as a reproducible platform for future research on active grasping. The source code of our paper will be made public when the paper is released to the public.

Adapting large language models (LLMs) to low-resource domains remains challenging due to the scarcity of domain-specific data. While in-domain data is limited, there exists a vast amount of general-domain data that shares similar question-answer formats and reasoning patterns with domain tasks. This observation raises an important question: can useful general-domain data be mined to improve low-resource domain adaptation? Our initial findings show that general-domain chain-of-thought data contains useful auxiliary signals for domain adaptation, even without careful selection. This observation motivates a new paradigm for domain adaptation beyond exclusive reliance on domain-specific data. To systematically identify the most beneficial general-domain samples, we propose NTK-Selector, motivated by the Neural Tangent Kernel's ability to capture alignment in training dynamics. Since directly applying NTK to pretrained LLMs is impractical, we introduce a Jacobian-free NTK approximation and empirically demonstrate stable NTK-like behavior during fine-tuning. Extensive experiments across medical, financial, legal, and psychological domains demonstrate that NTK-Selector consistently outperforms domain-only fine-tuning and existing data selection baselines. In particular, NTK-Selector achieves gains of +8.7 and +5.1 points on Llama3-8B-Instruct and Qwen3-8B, respectively, compared to only +0.8 and +0.9 points from domain-only fine-tuning.

While stereo matching has achieved maturity for 0D point and 1D line primitives, establishing correspondences for 2D polygons remains largely unexplored due to challenges including disparity discontinuity, scale variation, training dependency, and poor generalization, limiting downstream tasks such as pose estimation and 3D reconstruction. To address these issues, we are the first to propose a Zero-shot Polygon Matching paradigm with Pre-trained Models (i.e., Z(PM)2), which combines learned features and handcrafted geometric constraints through plug-and-play modules, extending matching from 0D/1D primitives to 2D polygons. The pipeline comprises three core stages: Firstly, detector leverages the pre-trained segment anything model to vectorize segmentation masks into graph-structured polygons integrating geometry and texture; Secondly, global matcher uses bidirectional-pyramid and multi-geometric constraints to handle viewpoint variation; Thirdly, local matcher leverages local-holistic bipartite graph optimization to resolve disparity discontinuity and topological inconsistency. Moreover, we develop polygon-matching-guided pose estimation using correspondences to obtain well-distributed, low-redundancy homologous points, and pioneer the polygon cloud concept with an optimal surface generation method, producing structurally complete and semantically rich 3D representations beyond point and line clouds. Since no polygon matching methods from stereo imagery are available for direct comparison, we selected state-of-the-art (SoTA) methods close to this task as baselines. Extensive experiments on five challenging datasets (ISPRS, KITTI, ScanNet, SceneFlow, DTU) show Z(PM)2 achieves a 68.60% matching area score, outperforming MESA by approximately 32% and ranking first in area-level pose estimation, with competitive speed and strong zero-shot generalization without any training requirement.

Deep reinforcement learning (DRL) has demonstrated remarkable success in developing autonomous driving policies. However, its vulnerability to adversarial attacks remains a critical barrier to real-world deployment. Although existing robust methods have achieved success, they still suffer from three key issues: (i) these methods are trained against myopic adversarial attacks, limiting their abilities to respond to more strategic threats, (ii) they have trouble causing truly safety-critical events (e.g., collisions), but instead often result in minor consequences, and (iii) these methods can introduce learning instability and policy drift during training due to the lack of robust constraints. To address these issues, we propose Intelligent General-sum Constrained Adversarial Reinforcement Learning (IGCARL), a novel robust autonomous driving approach that consists of a strategic targeted adversary and a robust driving agent. The strategic targeted adversary is designed to leverage the temporal decision-making capabilities of DRL to execute strategically coordinated multi-step attacks. In addition, it explicitly focuses on inducing safety-critical events by adopting a general-sum objective. The robust driving agent learns by interacting with the adversary to develop a robust autonomous driving policy against adversarial attacks. To ensure stable learning in adversarial environments and to mitigate policy drift caused by attacks, the agent is optimized under a constrained formulation. Extensive experiments show that IGCARL improves the success rate by at least 27.9% over state-of-the-art methods, demonstrating superior robustness to adversarial attacks and enhancing the safety and reliability of DRL-based autonomous driving.

Multi-page visual documents such as manuals, brochures, presentations, and posters convey key information through layout, colors, icons, and cross-slide references. While multimodal large language models (MLLMs) offer opportunities in document understanding, current systems struggle with complex, multi-page visual documents, particularly in fine-grained reasoning over elements and pages. We introduce SlideAgent, a versatile agentic framework for understanding multi-modal, multi-page, and multi-layout documents, especially slide decks. SlideAgent employs specialized agents and decomposes reasoning into three specialized levels--global, page, and element--to construct a structured, query-agnostic representation that captures both overarching themes and detailed visual or textual cues. During inference, SlideAgent selectively activates specialized agents for multi-level reasoning and integrates their outputs into coherent, context-aware answers. Extensive experiments show that SlideAgent significantly improves accuracy over both proprietary (+7.9%) and open-source models (+9.8%).

This paper investigates the limitations of the normalization in attention mechanisms. We begin with a theoretical framework that enables the identification of the model's selective ability and the geometric separation involved in token selection. Our analysis includes explicit bounds on distances and separation criteria for token vectors under softmax scaling. Through experiments with pre-trained GPT-2 model, we empirically validate our theoretical results and analyze key behaviors of the attention mechanism. Notably, we demonstrate that as the number of selected tokens increases, the model's ability to distinguish informative tokens declines, often converging toward a uniform selection pattern. We also show that gradient sensitivity under softmax normalization presents challenges during training, especially at low temperature settings. These findings advance current understanding of softmax-based attention mechanism and motivate the need for more robust normalization and selection strategies in future attention architectures.

Linear polymers, macromolecules formed from monomers covalently bonded into continuous chains, underpin countless technologies and are indispensable to modern life. While deep learning is advancing polymer science, existing methods typically represent the whole linear polymer solely through monomer-level descriptors, overlooking the global structural information inherent in polymer conformations, which ultimately limits their practical performance. Moreover, this important field still lacks a dedicated foundation model that can effectively support diverse downstream tasks, thereby severely constraining progress. To address these challenges, we introduce PolyConFM, a foundation model tailored for modeling and designing linear polymers through conformation-centric generative pretraining. Recognizing that each linear polymer is essentially a continuous chain whose conformation can be naturally decomposed into a sequence of local conformations (i.e., those of its repeating units), we pretrain PolyConFM under the conditional generation paradigm, reconstructing these local conformations via masked autoregressive (MAR) modeling and further generating their orientation transformations to recover the corresponding polymer conformation. Meanwhile, we construct a linear polymer conformation dataset via molecular dynamics simulations to mitigate data sparsity, thereby enabling conformation-centric pretraining. Experiments demonstrate that PolyConFM consistently outperforms representative task-specific methods across diverse downstream tasks, thereby equipping polymer science with a powerful tool targeting linear polymers.

At the core of Deep Research is knowledge mining, the task of extracting structured information from massive unstructured text in response to user instructions. Large language models (LLMs) excel at interpreting such instructions but are prohibitively expensive to deploy at scale, while traditional pipelines of classifiers and extractors remain efficient yet brittle and unable to generalize to new tasks. We introduce Falconer, a collaborative framework that combines the agentic reasoning of LLMs with lightweight proxy models for scalable knowledge mining. In Falconer, LLMs act as planners, decomposing user instructions into executable pipelines, and as annotators, generating supervision to train small proxies. The framework unifies classification and extraction into two atomic operations, get label and get span, enabling a single instruction-following model to replace multiple task-specific components. To evaluate the consistency between proxy models incubated by Falconer and annotations provided by humans and large models, we construct new benchmarks covering both planning and end-to-end execution. Experiments show that Falconer closely matches state-of-the-art LLMs in instruction-following accuracy while reducing inference cost by up to 90% and accelerating large-scale knowledge mining by more than 20x, offering an efficient and scalable foundation for Deep Research.

Autonomous robots often view rooms only partially, through a doorway, where the walls and scene structure hide the geometry and task-relevant semantics needed for safe navigation and goal-directed action. We ask whether off-the-shelf pretrained generative vision models can derive this missing structure as zero-shot offline priors for robot reasoning. Such priors should support spatio-semantic queries over unobserved structure, estimating the target object likelihood in hidden regions and the probability that those regions are occupied. Given an egocentric RGB observation and target query, our pipeline uses VLM-guided outpainting, monocular depth estimation, and semantic segmentation to sample semantically labeled 3D point cloud hypotheses of the hidden room. We introduce MatterDoor, a Matterport3D-derived benchmark of doorway-occluded indoor scenes, and evaluate the resulting priors with generative metrics and simulated Stretch robot object-reaching tasks. Our results suggest that useful spatio-semantic priors for planning can be derived without problem-specific fine-tuning.

In this paper, we propose a continuous-time formulation for the AdaGrad, RMSProp, and Adam optimization algorithms by modeling them as first-order integro-differential equations. We perform numerical simulations of these equations, along with stability and convergence analyses, to demonstrate their validity as accurate approximations of the original algorithms. Our results indicate a strong agreement between the behavior of the continuous-time models and the discrete implementations, thus providing a new perspective on the theoretical understanding of adaptive optimization methods.

Large Language Models (LLMs) have catalyzed vibe coding, where users leverage LLMs to generate and iteratively refine code through natural language interactions until it passes their vibe check. Vibe check reflects human preference and goes beyond functionality: the solution should feel right, read cleanly, preserve intent, and remain correct. However, current code evaluation remains anchored to pass@k and captures only functional correctness, overlooking non-functional instructions that users routinely apply. In this paper, we hypothesize that instruction following is the missing piece underlying vibe check besides functional correctness. To quantify models' code instruction-following capabilities with measurable signals, we present VeriCode, a taxonomy of 30 verifiable code instructions together with deterministic verifiers. We use the taxonomy to augment established evaluation suites, resulting in SWE-IF, a testbed to assess both instruction following and functional correctness. Evaluating 31 LLMs, we show that even the strongest models struggle to comply with multiple instructions and exhibit functional regression. Most importantly, a composite score of functional correctness and instruction following correlates best with human preference, with instruction following emerging as the primary differentiator among LLMs. Our code, data, and taxonomy are available at https://github.com/maszhongming/SWE-IF.

Adapting Foundation Models to new domains with limited training data is challenging and computationally expensive. While prior work has demonstrated the effectiveness of using domain-specific exemplars as in-context demonstrations, we investigate whether representing exemplars purely as text is the most efficient, effective, and stable approach. We explore an alternative: representing exemplars as soft prompts with an exemplar order invariant model architecture. To this end, we introduce Multi-Head Attention Retrieval-Augmented Generation (MHA-RAG), a framework with the number of attention heads serving as a simple hyperparameter to control soft prompt-generation across different tasks. Across multiple question-answering benchmarks and model scales, MHA-RAG achieves a 20-point performance gain over standard RAG, while cutting inference costs by a factor of 10X GFLOPs-delivering both higher accuracy and greater efficiency, invariant to exemplar order.

Recent advancements in generative models have allowed the emergence of a promising paradigm for recommender systems (RS), known as Generative Recommendation (GR), which tries to unify rich item semantics and collaborative filtering signals. One popular modern approach is to use semantic IDs (SIDs), which are discrete codes quantized from the embeddings of modality encoders (e.g., large language or vision models), to represent items in an autoregressive user interaction sequence modeling setup (henceforth, SID-based GR). While generative models in other domains exhibit well-established scaling laws, our work reveals that SID-based GR shows significant bottlenecks while scaling up the model. In particular, the performance of SID-based GR quickly saturates as we enlarge each component: the modality encoder, the quantization tokenizer, and the RS itself. In this work, we identify the limited capacity of SIDs to encode item semantic information as one of the fundamental bottlenecks. Motivated by this observation, as an initial effort to obtain GR models with better scaling behaviors, we revisit another GR paradigm that directly uses large language models (LLMs) as recommenders (henceforth, LLM-as-RS). Our experiments show that the LLM-as-RS paradigm has superior model scaling properties and achieves up to 20 percent improvement over the best achievable performance of SID-based GR through scaling. We also challenge the prevailing belief that LLMs struggle to capture collaborative filtering information, showing that their ability to model user-item interactions improves as LLMs scale up. Our analyses on both SID-based GR and LLMs across model sizes from 44M to 14B parameters underscore the intrinsic scaling limits of SID-based GR and position LLM-as-RS as a promising path toward foundation models for GR.

Multi-Agent Reinforcement Learning (MARL) provides a powerful framework for learning coordination in multi-agent systems. However, applying MARL to robotics remains challenging due to their high-dimensional continuous joint action spaces, complex reward design, and non-stationarity from concurrently learning agents. On the other hand, humans often learn complex coordination with the help of coaches, who guide learning through carefully designed curricula and detailed feedback. Building on the reasoning capabilities of foundation models, we argue that these models can similarly coach robots to learn coordination. Motivated by this, we propose CRAFT: Coaching Reinforcement learning Autonomously using Foundation models for learning coordination Tasks, a framework that leverages foundation models to act as a "coach" for multi-robot coordination. CRAFT automatically decomposes long-horizon coordination tasks into sequences of subtasks using the planning capability of Large Language Models (LLMs). Then, CRAFT trains each subtask using LLM-generated reward functions, and refines them through a Vision Language Model (VLM)-guided reward-refinement loop. We evaluate CRAFT on multi-quadruped navigation and bimanual manipulation tasks, and demonstrate its capability to learn complex coordination behaviors. In addition, in a multi-quadruped navigation setting, we show that our learned policies transfer to the real world. Project website is https://iconlab.negarmehr.com/CRAFT/

Scalability has driven recent advances in generative modeling, yet its principles remain underexplored for adversarial learning. We investigate the scalability of Generative Adversarial Networks (GANs) through two design choices that have proven to be effective in other types of generative models: training in a compact Variational Autoencoder latent space and adopting purely transformer-based generators and discriminators. Training in latent space enables efficient computation while preserving perceptual fidelity, and this efficiency pairs naturally with plain transformers, whose performance scales with computational budget. Building on these choices, we analyze failure modes that emerge when naively scaling GANs. Specifically, we find issues as underutilization of early layers in the generator and optimization instability as the network scales. Accordingly, we provide simple and scale-friendly solutions as lightweight intermediate supervision and width-aware learning-rate adjustment. Our experiments show that GAT, a purely transformer-based and latent-space GANs, can be easily trained reliably across a wide range of capacities (S through XL). Moreover, GAT-XL/2 achieves state-of-the-art single-step, class-conditional generation performance (FID of 2.18) on ImageNet-256 in just 60 epochs, 4x fewer epochs than strong baselines. Project page: https://hse1032.github.io/GAT.

Reasoning is a core capability of language models (LMs), yet it remains unclear how much model capacity is necessary to support reasoning during pretraining. In this work, we study the minimal parameter budget required for implicit reasoning, defined as the ability to infer new facts from learned knowledge without explicit chain-of-thought supervision. To isolate this phenomenon, we pretrain LMs from scratch in a controlled synthetic environment that mimics the structure and distribution of real-world knowledge graphs, and evaluate their ability to complete missing edges via multi-hop inference. From both a theoretical and an empirical perspective, we identify a scaling law linking this optimal parameter budget to a graph search entropy measure. Across a wide range of model sizes, training steps, and graph complexities, we show that an optimally sized language model can reliably reason over approximately 0.008 bits of information per parameter at most. Our results characterize the minimal sufficient capacity for implicit reasoning during pretraining. Our findings provide principled guidance for matching model size to data complexity and offer new insights into the scaling behavior of reasoning in large language models.

We introduce a sequence-conditioned critic for Soft Actor-Critic (SAC) that models trajectory context with a lightweight Transformer and trains on aggregated $N$-step targets. Unlike prior approaches that (i) score state-action pairs in isolation or (ii) rely on actor-side action chunking to handle long horizons, our method strengthens the critic itself by conditioning on short trajectory segments and integrating multi-step returns -- without importance sampling (IS). The resulting sequence-aware value estimates capture the critical temporal structure for extended-horizon and sparse-reward problems. On local-motion benchmarks, we further show that freezing critic parameters for several steps makes our update compatible with CrossQ's core idea, enabling stable training \emph{without} a target network. Despite its simplicity -- a 2-layer Transformer with 128-256 hidden units and a maximum update-to-data ratio (UTD) of $1$ -- the approach consistently outperforms standard SAC and strong off-policy baselines, with particularly large gains on long-trajectory control. These results highlight the value of sequence modeling and $N$-step bootstrapping on the critic side for long-horizon reinforcement learning.

Four-dimensional variational data assimilation (4DVAR) is a cornerstone of numerical weather prediction, yet it remains computationally intensive and sensitive to initialization due to the non-convexity of its objective function. We propose a neural field-based reformulation of 4DVAR in which the spatiotemporal state is represented as a continuous function parameterized by a neural network. We demonstrate that optimizing in parameter space leverages the spectral bias of neural fields, acting as an implicit regularizer that stabilizes state estimation and suppresses spurious high-frequency oscillations without requiring explicit background error covariance information. Furthermore, by parameterizing the full spatiotemporal trajectory, our framework enables parallel-in-time optimization and incorporates physical constraints directly through physics-informed losses. Evaluations on chaotic benchmarks, including 2D Kolmogorov flow and 3D Taylor-Green vortices, show that neural reparameterization produces more accurate initial conditions than classical 4DVAR. When combined with separable neural architectures (SPINNs), the method achieves substantial speedups. Unlike many machine learning approaches, this framework requires no ground-truth training data, offering a robust and scalable alternative for operational data assimilation.

This work proposes a framework LGKDE that learns kernel density estimation for graphs. The key challenge in graph density estimation lies in effectively capturing both structural patterns and semantic variations while maintaining theoretical guarantees. Combining graph kernels and kernel density estimation (KDE) is a standard approach to graph density estimation, but has unsatisfactory performance due to the handcrafted and fixed features of kernels. Our method LGKDE leverages graph neural networks to represent each graph as a discrete distribution and utilizes maximum mean discrepancy to learn the graph metric for multi-scale KDE, where all parameters are learned by maximizing the density of graphs relative to the density of their well-designed perturbed counterparts. The perturbations are conducted on both node features and graph spectra, which helps better characterize the boundary of normal density regions. Theoretically, we establish consistency and convergence guarantees for LGKDE, including bounds on the mean integrated squared error, robustness, and generalization. We validate LGKDE by demonstrating its effectiveness in recovering the underlying density of synthetic graph distributions and applying it to graph anomaly detection across diverse benchmark datasets. Extensive empirical evaluation shows that LGKDE demonstrates superior performance compared to state-of-the-art baselines on most benchmark datasets.

Autonomous stocking in retail environments, particularly supermarkets, presents challenges due to dynamic human interactions, constrained spaces, and diverse product geometries. This paper introduces an efficient modular robotic system for autonomous shelf stocking, integrating commercially available hardware with a scalable algorithmic architecture. A major contribution of this work is the system integration of off-the-shelf hardware and ROS2-based perception, planning, and control into a single deployable platform for retail environments. Our solution leverages Behavior Trees (BTs) for task planning, fine-tuned vision models for object detection, and a two-step Model Predictive Control (MPC) framework for precise shelf navigation using ArUco markers. Laboratory experiments replicating realistic supermarket conditions demonstrate reliable performance, achieving over 98% success in pick-and-place operations across a total of more than 700 stocking events. However, our comparative benchmarks indicate that the performance and cost-effectiveness of current autonomous systems remain inferior to that of human workers, which we use to highlight key improvement areas and quantify the progress still required before widespread commercial deployment can realistically be achieved.

A conventional LLM Unlearning setting consists of two subsets -"forget" and "retain", with the objectives of removing the undesired knowledge from the forget set while preserving the remaining knowledge from the retain. In privacy-focused unlearning research, a retain set is often further divided into neighbor sets, containing either directly or indirectly connected to the forget targets; and augmented by a general-knowledge set. A common practice in existing benchmarks is to employ only a single neighbor set, with general knowledge which fails to reflect the real-world data complexities and relationships. LLM Unlearning typically involves 1:1 sampling or cyclic iteration sampling. However, the efficacy and stability of these de facto standards have not been critically examined. In this study, we systematically evaluate these common practices. Our findings reveal that relying on a single neighbor set is suboptimal and that a standard sampling approach can obscure performance trade-offs. Based on this analysis, we propose and validate an initial set of best practices: (1) Incorporation of diverse neighbor sets to balance forget efficacy and model utility, (2) Standard 1:1 sampling methods are inefficient and yield poor results, (3) Our proposed Modular Entity-Level Unlearning (MELU) strategy as an alternative to cyclic sampling. We demonstrate that this modular approach, combined with robust algorithms, provides a clear and stable path towards effective unlearning.

The rapid expansion of interconnected devices, autonomous systems, and AI applications has created severe fragmentation in adaptive transport systems, where diverse protocols and context sources remain isolated. This survey provides the first systematic investigation of the Model Context Protocol (MCP) as a unifying paradigm, highlighting its ability to bridge protocol-level adaptation with context-aware decision making. Analyzing established literature, we show that existing efforts have implicitly converged toward MCP-like architectures, signaling a natural evolution from fragmented solutions to standardized integration frameworks. We propose a five-category taxonomy covering adaptive mechanisms, context-aware frameworks, unification models, integration strategies, and MCP-enabled architectures. Our findings reveal three key insights: traditional transport protocols have reached the limits of isolated adaptation, MCP's client-server and JSON-RPC structure enables semantic interoperability, and AI-driven transport demands integration paradigms uniquely suited to MCP. Finally, we present a research roadmap positioning MCP as a foundation for next-generation adaptive, context-aware, and intelligent transport infrastructures.