arXiv:2606.07928v2 Announce Type: replace Abstract: While previous versions of the Ai2 Climate Emulator (ACE) have been trained with CO$_2$ as a forcing, they are only accurate within a narrow range of scenarios, for example climate over the last 80 years forced by observed sea surface temperature (SST), sea ice, and CO$_2$ (AMIP), or equilibrium or near-equilibrium climates with CO$_2$ concentrations ranging from 1x to 4x that of the present day. Attempting to simulate climate forced by AMIP SST perturbed by +4 K or the response to an abrupt quadrupling of CO$_2$, results in unphysical behavior. We attribute this to these models being trained on datasets where the SST and CO$_2$ are correlated, limiting their ability to accurately learn their separate effects. In this study we introduce a new class of "random-CO$_2$" reference simulations where the SST and CO$_2$ are prescribed to vary independently. Trained on a balance of AMIP, equilibrium-climate, and random-CO$_2$ data, and including a total energy conservation constraint for improved interpretability, we present a more data-efficient model that not only accurately emulates its reference model in scenarios in which previous models excelled, but also scenarios like AMIP +4 K and slab-ocean-coupled abrupt 4xCO$_2$ where they did not. Limitations are that it has simplified or prescribed representations of other Earth system components like the ocean, land, and sea ice; does not expose other known climate drivers as forcings; and relies solely on physics-based model output for training data, inheriting the biases relative to observations thereof. Each of these represent opportunities for future work.
Science Journals
arXiv:2607.04108v1 Announce Type: new Abstract: Large language models are increasingly used as evolutionary engines for scientific discovery: generate candidates, select winners, feed them back as parents, and repeat. We audit whether this loop actually compounds discovery in scientific equation discovery, a setting where finite samples make structure underdetermined and interpolation easy. Under matched LLM-call budgets, parent-conditioned evolution is indistinguishable from fresh independent sampling: median OOD NMSE is 0.045 vs. 0.049, instructed multi-parent crossover is worse, final success is predicted by initial proposal quality, and multiple iteration schemes fail to add solved problems. Operationally, the loop reduces to what it produces: a dictionary of candidate terms. We turn that diagnosis into PTB-Search, a one-generation method for componentized scientific discovery. PTB-Search samples independent LLM proposals once, extracts reusable terms into a per-problem dictionary, and performs train-only set-level sparse selection with least-squares coefficients. Its central principle is that underdetermined data identifies the joint behavior of term sets, not reliable per-term credit. On identical dictionaries and zero additional LLM calls, set-level selectors solve 165--169 of 717 cells, while single-term reductions solve only 74--78. On the official 239-problem LLM-SRBench split, PTB-Search reaches 73.2% Acc0.1 with Llama-3.1-8B and 77.0% with a single-seed DeepSeek-V4 anchor, versus 49.2% for the best reported baseline, using one tenth of the standardized call budget. A program-domain stress test gives a scoped boundary: generation count remains unreliable, while retained external state can help in harder non-linear spaces. Across these results, LLMs are best understood as material suppliers; discovery is carried by external set-level selection over reusable components.
arXiv:2602.17395v2 Announce Type: replace Abstract: Generalized Category Discovery (GCD) aims to identify novel categories in unlabeled data while leveraging a small labeled subset of known classes. Training a parametric classifier solely on image features often leads to overfitting to old classes, and recent multimodal approaches improve performance by incorporating textual information. However, they treat modalities independently and incur high computational cost. We propose SpectralGCD, an efficient and effective multimodal approach to GCD that uses CLIP cross-modal image-concept similarities as a unified cross-modal representation. Each image is expressed as a mixture over semantic concepts from a large task-agnostic dictionary, which anchors learning to explicit semantics and reduces reliance on spurious visual cues. To maintain the semantic quality of representations learned by an efficient student, we introduce Spectral Filtering which exploits a cross-modal covariance matrix over the softmaxed similarities measured by a strong teacher model to automatically retain only relevant concepts from the dictionary. Forward and reverse knowledge distillation from the same teacher ensures that the cross-modal representations of the student remain both semantically sufficient and well-aligned. Across six benchmarks, SpectralGCD delivers accuracy comparable to or significantly superior to state-of-the-art methods at a fraction of the computational cost. The code is publicly available at: https://github.com/miccunifi/SpectralGCD.
arXiv:2602.20995v2 Announce Type: replace Abstract: Pre-ranking is a critical stage in industrial recommendation systems, tasked with efficiently scoring thousands of recalled items for downstream ranking. A key challenge is the train-serving discrepancy: pre-ranking models are trained only on exposed interactions, yet must score all recalled candidates -- including unexposed items -- during online serving. This mismatch not only induces severe sample selection bias but also degrades generalization, especially for long-tail content. Existing debiasing approaches typically rely on heuristics (e.g., negative sampling) or distillation from biased rankers, which either mislabel plausible unexposed items as negatives or propagate exposure bias into pseudo-labels. In this work, we propose Generative Pseudo-Labeling (GPL), a framework that leverages large language models (LLMs) to generate unbiased, content-aware pseudo-labels for unexposed items, explicitly aligning the training distribution with the online serving space. By offline generating user-specific interest anchors and matching them with candidates in a frozen semantic space, GPL provides high-quality supervision without adding online latency. Deployed in a large-scale production system, GPL improves click-through rate by 3.07%, while significantly enhancing recommendation diversity and long-tail item discovery.
arXiv:2607.04119v1 Announce Type: new Abstract: Reconstructing Computer-Aided Design (CAD) modeling sequences from images is crucial for preserving design intent and supporting parametric editing. However, existing methods typically generate full CAD sequences holistically, overlooking the iterative, feedback-driven nature of human design workflows. We address this limitation by introducing the rich stepwise visual supervision: at each modeling step, the system observes the target's orthographic projections, the projections of the incrementally constructed model, and the active sketch, enabling informed action selection. To effectively leverage this on-the-fly feedback, we propose SOV-CAD, a framework that formulates CAD reconstruction as a sequential decision-making task and employs offline reinforcement learning with a Decision Transformer architecture. This design incorporates continuous visual feedback guided by geometric alignment rewards, resulting in a more accurate and human-like modeling process. Extensive experiments show that SOV-CAD surpasses state-of-the-art methods in CAD sequence reconstruction while exhibiting strong data efficiency. Code of SOV-CAD is available at: https://github.com/LukePhong/SOV-CAD
arXiv:2607.04127v1 Announce Type: new Abstract: We present a real-time LiDAR-based framework for Gaussian Splatting SLAM that tightly couples fast G-ICP registration with spherical rasterization-based dense mapping for large-scale sequences. Leveraging LiDAR geometry rather than appearance, we reuse tracking-estimated local covariances to initialize Gaussians with range-aware scales and to derive surface normals for geometry-aware map optimization. We further introduce a covariance-derived geometry score that measures local complexity and drives pruning in planar regions and selective densification in structurally rich areas, while optimized Gaussians and LiDAR-specific confidence cues are fed back to improve tracking robustness. On the Newer College dataset, our method achieves an F-score of 86.78\% using purely online trajectories at real-time speed ($>$20 FPS), and additional experiments on other datasets confirm its stability and scalability.
arXiv:2607.04388v1 Announce Type: new Abstract: We provide an online algorithm with the following guarantee: for any fixed sequence of vectors $v_1,\dots,v_T \in \mathbf{R}^d$ with $\|v_i\|_2\le 1$, the algorithm assigns each arriving vector $v_t$ a random sign $\varepsilon_t$ such that every prefix sum $\sum_{i=1}^t \varepsilon_i v_i $ can be written as the sum of three coupled standard Gaussian vectors. Our algorithm runs in $O(dT)$ time and achieves the optimal prefix discrepancy bound \[ \max_{1 \le t \le T}\left\| \sum_{i=1}^t \varepsilon_i v_i \right\|_\infty = O\left( \sqrt{\log T} \right), \] with high probability. This recovers the optimal bound of Kulkarni, Reis, and Rothvoss, whose algorithm runs in time exponential in $T$ and $d$. The algorithm and main proof were discovered in a GPT-5.5 Pro Extended conversation prompted by the author.
arXiv:2603.05169v3 Announce Type: replace Abstract: Local energy markets empower prosumers in distribution grids to form coalitions for collective self-consumption. An open question is to analyze the scale and composition of local energy market coalitions formed by strategic prosumers in distribution grids. This analysis must account for grid constraints, stochasticity of load and generation, as well as the interaction between possibly multiple local energy markets in the distribution grid. In this work, we present a cooperative game theoretic framework to study distribution grid partitioning into local energy markets under uncertain prosumption, grid constraints, and coalitional externalities. We formulate the optimal stable partitioning problem to balance the interests of the grid operator with that of strategic prosumers. Under deterministic load and generation, we show that the largest market coalition is the optimal stable partition. Under high levels of grid congestion, we show that individual self-consumption corresponds to the optimal stable partition. For the general case of stochastic prosumption and moderate grid congestion levels, we provide an algorithm to evaluate the optimal stable partition. We validate our algorithm and theory using numerical experiments on benchmark and real world distribution grids. Our results help in understanding the impact of prosumption uncertainty and grid constraints on coalition formation.
arXiv:2602.13517v2 Announce Type: replace Abstract: Large language models (LLMs) have demonstrated impressive reasoning capabilities by scaling test-time compute via long Chain-of-Thought (CoT). However, recent findings suggest that raw token counts are unreliable proxies for reasoning quality: increased generation length does not consistently correlate with accuracy and may instead signal "overthinking," leading to performance degradation. In this work, we quantify inference-time effort by identifying deep-thinking tokens -- tokens where internal predictions undergo significant revisions in deeper model layers prior to convergence. Across four challenging mathematical and scientific benchmarks (AIME 24/25, HMMT 25, and GPQA-diamond) and a diverse set of reasoning-focused models (GPT-OSS, DeepSeek-R1, and Qwen3), we show that deep-thinking ratio (the proportion of deep-thinking tokens in a generated sequence) exhibits a robust and consistently positive correlation with accuracy, substantially outperforming both length-based and confidence-based baselines. Leveraging this insight, we introduce Think@n, a test-time scaling strategy that prioritizes samples with high deep-thinking ratios. We demonstrate that Think@n matches or exceeds standard self-consistency performance while significantly reducing inference costs by enabling the early rejection of unpromising generations based on short prefixes.
arXiv:2606.07994v2 Announce Type: replace Abstract: Scientific progress is traditionally narrated through the interplay of theoretical insights and experimental findings. Yet this view of science underplays a third and central pillar of progress: the methods that underlie both conceptual advances and empirical evidence. By analysing more than 3 million articles across science published between 1980 and 2019, we find that science has undergone a fundamental structural transition. The share of papers that primarily contribute new methods-methods papers-has doubled across science over the past four decades, rising universally across disciplines and citation impact levels. Rather than a gradual evolution, this transition marks a pivotal shift beginning in the early 1990s, aligning with the computational revolution and the emergence of data-intensive science. The surge in methodological research is not confined to the most cited, elite publications; it spans the full spectrum of scientific output. These findings reveal a systemic reorientation of the scientific ecosystem where reusable methods increasingly serve as the essential infrastructure of scientific advances, challenging the traditional dichotomy of theory and experimental research. As science becomes increasingly methods-driven, our results call for rethinking how research is evaluated, funded and organised-towards better incentivising method innovations. This is especially the case as expanding AI must be effectively integrated with scientific instruments to realise its full potential.
arXiv:2606.08281v2 Announce Type: replace Abstract: Safe physical human-robot interaction (pHRI) is fundamentally a problem of interaction dynamics: the robot must track a commanded motion, yield under human forces, respect actuator and joint limits, and stay predictable under persistent contact. Classical impedance control shapes this through a virtual spring-damper, but a sustained force produces the bias $e_\infty=-K_d^{-1}F_h$, trading accuracy for safety. We propose a predictive framework that makes interaction dynamics explicit through a linear double-integrator backbone: an operational-space feedforward cancels gravity and Coriolis terms and normalizes the task inertia, leaving a configuration-independent state-transition matrix with robot dependence isolated in the input matrix. This converts nonlinear torque-controlled pHRI into a linear constrained-control problem, so offset-free tracking, actuator feasibility, sampled-data joint-limit safety, and passivity filtering follow with explicit assumptions. The online realization is a 30-variable convex QP at 100 Hz with a precomputed free-response matrix and a Kalman filter that rejects persistent forces without steady-state error; null-space barrier, one-step joint-limit CBF, and energy-tank filters add conditional safety and task-channel passivity. In MuJoCo simulation of a 7-DOF Franka FR3, the controller attains sub-0.05 mm steady-state error under a sustained 15 N force versus 44.8 mm for classical impedance, sub-millimeter tracking on four 3-D circles, and robustness to measurement noise and 30% inertial mismatch.
arXiv:2606.08362v2 Announce Type: replace Abstract: Existing scientific relation extraction benchmarks mainly target domains such as computer science, where entities are tasks, methods, datasets, materials, or metrics. This leaves a gap in variable-oriented empirical fields such as psychology, where findings are expressed as relations among constructs, measurements, interventions, and outcomes. We introduce variable-centered empirical graph extraction, the task of mapping scientific abstracts to typed graphs whose nodes are normalized variables and whose edges represent empirical and hierarchical relations. To support this task, we construct EmpiriGraph-Psy, a benchmark of 210 psychology abstracts annotated by domain-trained annotators with normalized variables, concept hierarchies, empirical relation types, and validation states. We evaluate frontier and open-weight LLMs using both direct extraction and a staged graph-construction pipeline that separates variable extraction, normalization, hierarchy construction, evidence selection, relation extraction, and edge validation. The staged pipeline substantially outperforms direct extraction, with the best configuration achieving a macro-F1 of 0.74. Error analysis shows that moderation relations and concept hierarchies remain the most challenging cases, highlighting the difficulty of extracting higher-order empirical claims and implicit abstraction structure from scientific abstracts.
arXiv:2606.09118v3 Announce Type: replace Abstract: Evaluation protocols can lag behind LLM capabilities. Programmatically verified benchmarks cover narrow surface constraints, whereas real-world instruction following and agentic workflows require judging semantic, contextual, and policy-dependent behavior. We study expert-curated rubric-based evaluation as a unified mechanism for measurement and reinforcement-learning rewards across two settings: complex instruction following and enterprise agentic tasks. We identify rubric-design choices that affect reward quality, including maximum viable atomicity, intent-aware criterion design, and LLM-judge calibration. We introduce ComplexConstraints, an expert-curated instruction-following suite comprising a public 75-prompt benchmark with 1,559 rubric criteria and a disjoint 1,000-prompt training set, with 10-40 atomic criteria per prompt. Empirically, rubric rewards improve training in both fixed task datasets, such as ComplexConstraints, and stateful RL environments, such as CoreCraft. Training a 4B model on ComplexConstraints improves mean criterion pass rate by +15.5 pp on a held-out split, bringing it within 0.5 pp of the untrained baseline of a roughly 60x larger Qwen3 model, and the gains transfer to external benchmarks the model never saw during training: +8.4 pp on AdvancedIF and +10.1 pp on MultiChallenge. In CoreCraft, rubric-reward RL likewise transfers to out-of-distribution benchmarks (+4.5 pp BFCL, +7.4 pp tau^2-Bench, +6.8 pp Toolathlon). These results show that expert-authored rubrics provide effective evaluation targets and scalable reward signals for improving LLM instruction following and agentic behavior.
arXiv:2606.09209v5 Announce Type: replace Abstract: Frequent Itemset Mining (FIM) is an important task in data analytics, where classical algorithms face scalability bottlenecks from the combinatorial growth of candidates and the memory overhead of their data structures. Inspired by recent developments in quantum computing, in this paper, we propose the Quantum Frequent-itemset Mining (QFM) data-processing framework for FIM. Following the level-wise structure of the itemset lattice, QFM introduces three mechanisms: (1) Bit-Vector Qubit Encoding for quantum data representation, which organizes transaction data into branchless bit-vectors to facilitate systematic uncomputation; (2) Mining-Aware Candidate Superposition, which prepares a quantum superposition over valid candidates at each lattice level rather than the full itemset lattice; and (3) Bit-Parallel Threshold Marking, which constructs a logarithmic-depth threshold-marking oracle for reliable repeated support verification within hardware coherence limits. We provide theoretical time complexity analysis, implement QFM on IBM Qiskit and Amazon Braket, and evaluate it on real-world datasets against representative classical baselines, where QFM achieves 96% improvement on average.
arXiv:2606.09426v3 Announce Type: replace Abstract: Computer-use agents (CUAs) increasingly operate in runtimes that combine visual desktop control, command-line execution, code editing, browsers, and external tools. Existing benchmarks, however, often evaluate these interfaces as separable capabilities, leaving long-horizon cross-interface orchestration under-tested. Thus, we introduce WeaveBench, a long-horizon hybrid-interface benchmark with 114 tasks across 8 real-world work domains, grounded in real user requests and publicly verifiable artifacts. Each task requires agents to combine GUI observations/actions with CLI/code operations within a single trajectory. We evaluate these tasks on a real Ubuntu desktop inside deployed CLI-agent runtimes, augmented with a minimal desktop-control plugin. We also propose a companion trajectory-aware judge that inspects deliverables, files, screenshots, logs, and action traces, while detecting shortcut behaviors such as fabricated visual evidence or hard-coded metrics. Across frontier model-runtime pairings, the best PassRate reaches only 41.2%, showing the benchmark remains far from saturated. The trajectory-aware judge further reveals that outcome-only grading substantially overestimates agent performance. Overall, WeaveBench exposes a critical gap in CUA evaluation and provides an effective testbed to measure whether agents can orchestrate GUI, CLI, and code operations across long-horizon real-world tasks.
arXiv:2606.10451v2 Announce Type: replace Abstract: We study optimal pricing of versioned data products when buyers can combine multiple purchases. A monopoly seller offers a menu of data products, and a buyer's value for data is the improvement to their expected utility in a Bayesian decision problem. Since a buyer may purchase any finite bundle of products, including repeated copies of the same product, versioning creates arbitrage opportunities: a bundle of cheaper products may be more valuable than a product with a higher price. We formulate the arbitrage-free data selling problem which has infinite arbitrage-free constraints in general and possibly infinite state space, and prove its computational intractability: the problem admits no PTAS even for instances in which the state space is finite, and the problem admits no polynomial time constant factor approximation for succinct high-dimensional instances. On the positive side, when the numbers of buyer types and actions are constant, we give an additive FPTAS that handles possibly infinite state spaces and infinitely many arbitrage-free constraints, and empirically validate the algorithm on realistic synthetic data trading scenarios. We also analyze the posted pricing algorithm for selling only complete information and prove a tight approximation factor. We further identify a threshold utility regime in which arbitrage-freeness reduces to Blackwell dominance, which unifies known arbitrage-free conditions for dataset query and machine learning model pricing. Under this regime, we design efficient algorithms for several structured data menus common in practice.
arXiv:2606.10522v2 Announce Type: replace Abstract: Graphical User Interfaces (GUIs) serve as the dominant medium for human-computer interaction, yet building GUI agents that generalize across the vast diversity of real-world interface environments, with the same flexibility and robustness that humans naturally exhibit, remains unsolved. Notably, GUI data are inherently non-stationary: the continual emergence of previously unseen interface instances (e.g., novel domains and resolutions) induces persistent distribution shifts, significantly impeding the continual learning of existing GUI agents. Reinforcement fine-tuning (RFT) has attracted considerable attention as a promising approach. Nevertheless, RFT exhibits pronounced instability in its grounding capability, manifested as sharp reward discontinuities and high-variance oscillations. The imbalanced distribution of rollout outcomes introduces substantial noise into advantage estimation, leading to policy overconfidence. The fixed clipping bound suppresses the increase in policy probabilities needed to adapt to new distributions, leading to a collapse in exploration capacity. To address these challenges, we propose GUI-AC, a method that enhances the continual learning capability of GUI agents. GUI-AC introduces grounding certainty to support two core mechanisms: (i) Adaptive Advantage, which down-weights noisy advantage estimates to prevent policy overconfidence; and (ii) Dynamic Clipping, which relaxes the clipping bound to encourage exploration range. Extensive experiments show that these mechanisms jointly improve performance, enabling our method to surpass state-of-the-art baselines. Code is available anonymously at https://github.com/Can-Lin/GUI-AC.
arXiv:2407.11691v5 Announce Type: replace Abstract: We present VLMEvalKit: an open-source toolkit for evaluating large multi-modality models based on PyTorch. The toolkit aims to provide a user-friendly and comprehensive framework for researchers and developers to evaluate existing multi-modality models and publish \textbf{reproducible} evaluation results. In VLMEvalKit, we implement over 450+ large multi-modality model configurations, including both proprietary APIs and open-source models, and support 330+ benchmarks across diverse multi-modal benchmarks. By implementing a single interface, new models can be easily added to the toolkit, while the toolkit automatically handles the remaining workloads, including data preparation, distributed inference, prediction post-processing, and metric calculation. VLMEvalKit has also evolved to a broader evaluation suite spanning video/audio, document understanding, GUI grounding, spatial reasoning, safety, scientific reasoning, and multi-turn dialogue. Based on the evaluation results obtained with the toolkit, we host the OpenVLM Leaderboard, a comprehensive leaderboard to track the progress of multi-modality learning research. The toolkit is released on https://github.com/open-compass/VLMEvalKit and is actively maintained.
arXiv:2607.04647v1 Announce Type: cross Abstract: Scalable Bayesian inference for generalized linear mixed models (GLMMs) provides uncertainty-aware analysis of correlated longitudinal data, but existing scalable approaches largely assume low-dimensional tabular predictors and do not directly accommodate high-dimensional modalities such as images and text. We address this limitation by learning one or more modality-specific neural encoders jointly with a GLMM objective, then performing variance-corrected stochasticgradient MCMC for the GLMM parameters conditional on the learned representation. This conditional-Bayes design combines supervised representation learning with posterior uncertainty quantification for population-level effects, subjectspecific heterogeneity, and modality-level random slopes. The resulting model preserves interpretable fixed and random effects for structured covariates and learned modalities while scaling gracefully to large longitudinal datasets. In simulation studies, our method recovers posterior means and variance estimates from full-data MCMC benchmarks after covariance correction. We further evaluate uncertainty through parameter-level interval coverage in simulations and predictive calibration on held-out data. Applications to glaucoma progression and adolescent mental health demonstrate that the framework allows nuanced assessment of the relative importance of each modality on both individual and population levels without sacrificing predictive performance.
arXiv:2607.04197v1 Announce Type: cross Abstract: We have installed and operated a plastic scintillator detector counter to measure the flux of cosmic radiation during the 2023-2025 tour of the historical vessel Amerigo Vespucci. The Vespucci is the oldest ship of the Italian Navy and serves as a training vessel for Navy cadets. During its tour, some experiments were hosted onboard the vessel, providing unique opportunities for scientists working in different fields. We installed our detector upon the Vespucci's departure from Darwin in early October 2024. The detector collected cosmic-ray data during the journey from Darwin to Trieste, where the worldwide tour ended in March 2025. After about one month of stop in Trieste, the ship continued its tour in the Mediterranean sea, and arrived in Genova on June 10, 2025. We performed measurements of the cosmic radiation reaching the sea level across a wide latitude range, from 15{\deg} S to about 45{\deg} N. The lowest rate (averaged over all azimuth angles) was measured at a geographic latitude of about 7{\deg} N, and was about 16% less than the highest value, which was measured at Trieste, the northernmost location of the journey. Latitude effects on the cosmic radiation flux at sea level are due to the quasi-dipole geomagnetic field configuration, tilted by an angle of about 11{\deg} with respect to Earth's rotational axis.
arXiv:2503.17577v2 Announce Type: replace Abstract: Deepfakes have emerged as a widespread and rapidly escalating concern in generative AI, spanning images, audio, and videos. Among these, audio deepfakes are particularly alarming due to the growing accessibility of high-quality voice synthesis tools and the ease with which synthetic speech can be distributed through social media and robocalls. Consequently, detecting audio deepfakes is critical for combating the misuse of AI-generated speech. However, real-world audio is often affected by corruptions such as noise, audio modification, and compression, which can significantly degrade detection performance. In this work, we systematically evaluate the robustness of 10 audio deepfake detection models against 18 common corruption types, grouped into three categories: noise perturbation, audio modification, and compression. Using both traditional deep learning models and state-of-the-art speech foundation models, our study yields four key insights. (1) Most models are robust to noise but remain vulnerable to audio modifications and compression, especially neural codecs. (2) Speech foundation models consistently outperform traditional models across most corruption scenarios, likely due to large-scale pre-training on diverse audio datasets. (3) Increasing model size improves robustness, although the gains diminish as models become larger. (4) Robustness to unseen corruptions can be improved through targeted data augmentation during training or speech enhancement at inference time. These findings highlight the importance of evaluating audio deepfake detectors under diverse real-world corruptions and developing more robust detection frameworks for practical deployment. We further advocate that future research on deepfake detection across all media should account for the diverse and unpredictable distortions encountered in real-world environments.
arXiv:2503.19631v4 Announce Type: replace Abstract: We study applications of clustering (in particular, the $k$-center clustering problem) in the design of efficient and practical algorithms for computing an approximate and the exact arithmetic matrix product of two 0-1 rectangular matrices with clustered rows or columns, respectively. Our results in part can be regarded as an extension of the clustering-based approach to Boolean square matrix multiplication due to Arslan and Chidri (CSC 2011). First, we provide a simple and efficient deterministic algorithm for approximate matrix product of 0-1 matrices, where the additive error is proportional to the minimum maximum radius in an $\ell$-center clustering of the rows of the first matrix or an $k$-center clustering of the columns of the second matrix. Next, we use the approximation algorithm as a preprocessing after which a query asking for the exact value of an arbitrary entry in the product matrix can be answered in time proportional to the additive error. As a consequence, we obtain a simple deterministic algorithm for the exact matrix product of 0-1 matrices. We also present an improved simple deterministic algorithm for the exact product and in addition, faster analogous randomized algorithms for an approximate and the exact matrix products of 0-1 matrices based on randomized $\ell$ and $k$-center clustering.
arXiv:2607.04661v1 Announce Type: new Abstract: Reconstructing 3D scene structures from sparse, low-overlap observations remains a fundamental challenge in autonomous driving. Recent state-of-the-art frameworks achieve promising results by incorporating voxel-based Gaussians, but incur substantial computational redundancy due to a uniform volumetric processing strategy. To bridge the gap between the efficiency of pixel-based Gaussian methods and the structural completeness of voxel-based Gaussian approaches, we propose FocusGS, a simple yet effective framework that shifts the paradigm from global densification to targeted structural completion. Our central insight is that structural completion should be decoupled from deterministic regions, with computation concentrated exclusively on areas exhibiting geometric ambiguity. Specifically, FocusGS addresses the localization challenge by deriving a 3D Geometric Ambiguity Manifold to accurately isolate localized areas prone to occlusion and high geometric uncertainty. To overcome the subsequent manifold completion challenge, we design a lightweight targeted structure completion module that selectively instantiates and optimizes continuous Gaussian queries strictly within this unstructured, sparse topological subspace. Extensive experiments demonstrate that FocusGS achieves a superior efficiency-quality trade-off, advancing state-of-the-art performance on driving-centric benchmarks while naturally reducing the total number of Gaussians by ~74% and decreasing rendering time by ~34%.
arXiv:2607.04681v1 Announce Type: new Abstract: Embodied Chain-of-Thought has emerged as a promising mechanism to enhance robot decision-making and interpretability in black-box Vision-Language Action (VLA) models. However, whether this verbalized Chain-of-Thought truthfully reflects the policy's underlying decision process remains poorly understood. We distinguish between functional reasoning, in which reasoning improves task performance, and faithful reasoning, in which reasoning truly reflects the policy's internal decision process. We argue that SoTA alignment strategies offer a necessary but insufficient notion of faithfulness, admitting reasoning whose intermediate steps can mask the causal links in action prediction through confounding factors (e.g., reasoning that is ungrounded in the environment and internally disconnected or inconsistent), restricting policy generalization. We study this gap through a human evaluation of a SoTA reasoning model for autonomous driving, revealing an inconsistent coupling between reasoning quality and downstream trajectory improvement. We then operationalize a behavioral surrogate for embodied faithfulness through a learned critic, Pinocchio, scoring observation grounding and stepwise coherence, and use this critic as a dense reward signal in post-training an embodied policy with reinforcement learning. Across withheld driving benchmarks, our post-trained planner improves faithfulness by 4% and 18% over SoTA alignment and trajectory error post-training baselines, respectively, while maintaining competitive downstream task performance. Finally, on a synthetic out-of-distribution test set, post-training for faithfulness improves policy responsiveness to rare counterfactual scenarios by 1.6x that of a SoTA policy, suggesting that faithful reasoning traces contribute to more robust, generalizable, and interpretable embodied intelligence. Project page: https://mjf-su.github.io/pinocchio/
arXiv:2605.01220v2 Announce Type: replace Abstract: Visual Autoregressive Modeling (VAR) based on next-scale prediction achieves strong generation quality, but their explicit deep stacks fix the amount of computation per scale and inflate memory at high resolutions. We introduce Visual Implicit Autoregressive Modeling (VIAR), a next-scale autoregressive generator that embeds an implicit equilibrium layer between shallow pre/post blocks. The implicit layer is trained with Jacobian-Free Backpropagation, yielding constant training memory, while inference exposes a per-scale iteration knob that enables compute control. On ImageNet 256x256 benchmark, VIAR attains FID 2.16, and sFID 8.07 with only 38.4% parameters of VAR, matching or surpassing strong AR baselines and remaining competitive with large diffusion models. By controlling the per-scale knob, VIAR can reduce peak memory from 19.24 GB to 8.53 GB and doubles throughput from 15.16 to 32.08 images/s on a single RTX 4090, without retraining. Ablations show that fewer steps are sufficient for fixed-point iterations to converge and that VIAR consistently dominates VAR across quality efficiency operating points. In zero shot in-painting and class-conditional editing, VIAR produces sharper details and smoother boundaries while preserving global structure, validating the benefits of implicit equilibria and per-scale compute control for practical, deployable visual generation.