arXiv:2605.19485v1 Announce Type: new
Abstract: Large Reasoning Models (LRMs) have demonstrated remarkable capabilities in solving complex problems by generating structured, step-by-step reasoning content. However, exposing a model's internal reasoning process introduces additional safety risks; for example, recent studies show that LRMs are more vulnerable to jailbreak attacks than standard LLMs. In this paper, we investigate jailbreak attacks on LRMs and reveal that the attack success rate (ASR) is closely correlated with LRMs' attention patterns. Specifically, successful jailbreaks tend to assign lower attention to harmful tokens in the input prompt, while allocating higher attention to those tokens in the reasoning content. Motivated by this finding, we propose a novel jailbreak method for LRMs that leverages reinforcement learning (RL) to enhance attack effectiveness, explicitly incorporating attention signals into the reward function design. In addition, we introduce diverse persuasion strategies to enrich the RL action space, which consistently improves the ASR. Extensive experiments on five open-source and closed-source LRMs across three benchmarks demonstrate that our method achieves substantially higher ASR, outperforming existing approaches in terms of effectiveness, efficiency, and transferability.
Science Journals
arXiv:2605.18828v1 Announce Type: new
Abstract: The purpose of this style guide is to provide a clear, consistent framework for the design and development of human system interfaces (HSIs) used throughout the Fermilab accelerator complex. It establishes a shared visual and interaction foundation to ensure that interfaces remain intuitive, effective, and cohesive, regardless of when or by whom they are developed. By adhering to these guidelines, developers can avoid introducing unnecessary deviations that compromise usability or increase system training burden. This consistency is especially critical in long-term, multi-contributor projects where interface continuity and maintainability are paramount. This document serves as a practical reference for all HSI development activities related to the accelerator control environment. While the guidance provided is comprehensive, it is not exhaustive of every potential design scenario. As such, the style guide is intended to function as a living document, subject to regular review and revision. Updates will be made at least annually to incorporate emerging best practices, operational feedback, and evolving system needs. Areas where detailed guidance is still under development are clearly indicated in gray throughout the document and will be addressed in future revisions according to project priorities.
arXiv:2503.08633v2 Announce Type: replace
Abstract: Machine unlearning is the task of updating a trained model to forget specific training data without retraining from scratch. In this paper, we investigate how unlearning of deep neural networks (DNNs) is affected by the model parameterization level, which corresponds here to the DNN width. We define validation-based tuning for several unlearning methods from the recent literature, and show how these methods perform differently depending on (i) the DNN parameterization level, (ii) the unlearning goal (unlearned data privacy or bias removal), (iii) whether the unlearning method explicitly uses the unlearned examples. Our results show that unlearning usually excels on overparameterized models by significantly improving privacy/bias at a reasonable cost of utility (generalization) degradation; although for bias removal this requires the unlearning method to use the unlearned examples. Furthermore, we measure how much the unlearning changes the classification decision regions in the proximity of the unlearned examples, and avoids changing them elsewhere. By this we show that the unlearning success for overparameterized models stems from the ability to delicately change the model functionality in small regions in the input space while keeping much of the model functionality unchanged.
arXiv:2605.18824v1 Announce Type: new
Abstract: Evaluation of foundation models often rely on aggregate scores from benchmarks that lack comprehensive coverage and metadata for a fine-grained evaluation. We introduce a framework for automated benchmark generation. Our framework generates evaluation problems grounded in reference material, such as textbooks, producing benchmarks with broad coverage, rich metadata, and robustness to contamination. The pipeline employs a multi-agent architecture for problem generation and a solution-graph-driven strategy that significantly improves the reliability of ground truth solutions. Using the framework, we generate three benchmarks in Machine Learning, Corporate Finance, and Personal Finance. Expert review finds a significantly lower ground-truth error rate than previous benchmarks such as MMLU and GSM8K. Evaluation of 12 commercial and open-source models shows that our benchmarks achieve near-uniform competency coverage and surface performance differences across models that existing benchmarks fail to capture. We will open-source the framework and our curated benchmarks soon.
arXiv:2605.18822v1 Announce Type: new
Abstract: Post-training has become essential for adapting large language models (LLMs) to complex downstream behaviors, including instruction following, preference alignment, and multi-step reasoning. Reinforcement learning with verifiable rewards (RLVR) has recently emerged as a particularly effective post-training paradigm for improving reasoning capabilities, with critic-free algorithms such as GRPO and GSPO enabling scalable optimization. However, RLVR post-training with full fine-tuning (FFT) requires substantial GPU memory and incurs high training costs. Although parameter-efficient fine-tuning (PEFT) methods, such as Low-Rank Adaptation (LoRA), effectively reduce computational costs, they often suffer from a noticeable performance gap compared to full fine-tuning in post-training for complex reasoning tasks. In this paper, we propose Hybrid-LoRA, an efficient hybrid post-training framework that selectively applies full fine-tuning to a small subset of modules less suited to low-rank adaptation, while adapting the remaining components with LoRA. We introduce a novel Hybrid-LoRA Score to rank candidate modules according to their sensitivity to low-rank adaptation under a fixed parameter budget. Experiments show that Hybrid-LoRA closely matches full fine-tuning performance under a 10% full fine-tuning module budget, with the remaining candidate modules adapted by LoRA, consistently outperforming four state-of-the-art PEFT post-training baselines, achieving improvements of up to 5.65% and on average 4.36% over the best baseline.
arXiv:2605.18816v1 Announce Type: new
Abstract: Neural surrogates enable orders-of-magnitude acceleration of computational fluid dynamics (CFD) simulations, with the potential to transform engineering and healthcare workflows. Neural surrogate use in real-world applications requires addressing scalability to large, high-resolution surface and volume meshes, as well as to bespoke architectures, and accounting for limited training data through the use of inductive biases. Group-equivariant architectures are a principled way to introduce such bias, yet they can be detrimental when the learning problem itself breaks symmetry, for example, due to strong distributional alignment in the dataset. In this work, we investigate under which conditions equivariance improves generalization in neural CFD surrogates across tasks with increasing levels of distributional alignment and realism, covering automotive aerodynamics and blood flow (hemodynamics). To systematically assess the added value of equivariance at the limit of problem scaling, we introduce the Anchored-Branched Geometric Algebra Transformer (AB-GATr), a neural surrogate that integrates scalability and symmetry preservation to efficiently model coupled surface and volume quantities in an $E(3)$-equivariant manner. We find that on strongly aligned aerodynamics datasets, i.e., those that break symmetry, enforcing equivariance can degrade in-distribution performance. In contrast, across hemodynamic benchmarks with diverse geometries and varying alignment, equivariance is consistently beneficial. Moreover, across all benchmarks, the explicit equivariance of AB-GATr reliably outperforms implicit symmetry learning through data augmentation. Our findings showcase that equivariance is not universally beneficial across domains, yet it brings tangible advantages in problems lacking strong data regularities.
arXiv:2605.19466v1 Announce Type: new
Abstract: We report precision spectroscopy of ionic-core transitions in alkaline-earth Rydberg atoms. We demonstrate high-resolution measurements of isotope shifts and hyperfine splitting of dipole transitions in ionic cores which have not been explored so far. A key element of this work is the reduction of the linewidth by more than two orders of magnitude enabled by dynamical control of Rydberg electron's orbit which significantly enhances the spectral resolution. Furthermore, to unambiguously identify the frequency shift, we directly compare core ion's spectrum with a signal from a single trapped ion serving as an ultimate frequency reference. This work provides an important foundation for quantum control of inner-core transitions, which offer an useful tool in manipulating Rydberg atom as well as a sensitive probe for electron-core interactions in atomic and molecular systems.
arXiv:2602.03139v2 Announce Type: replace
Abstract: Distribution matching distillation (DMD) facilitates few-step image generation by aligning a distilled student with a reference multi-step teacher. In practice, however, optimizing DMD can reduce sample diversity in few-step synthesis, and existing remedies typically rely on perceptual or adversarial regularization, leading to stability and scalability challenges during training. Here, we describe diversity-preserved DMD (DP-DMD), a role-separated distillation method inspired by the complementary roles of early and late denoising steps. Specifically, the first distillation step is trained with a teacher-derived target-prediction objective (e.g., v-prediction) to preserve sample diversity, while the remaining steps are optimized with the standard DMD loss to refine perceptual quality. DP-DMD, with no perceptual or adversarial regularization, no additional modules, and no teacher-generated reference samples, preserves sample diversity while maintaining competitive visual quality under few-step sampling, providing a simple and stable alternative to other DMD variants.
arXiv:2605.19621v1 Announce Type: cross
Abstract: Deep generative models have emerged as state-of-the-art for solving inverse problems, but applying them to inverse problems for PDEs, like electrical impedance tomography (EIT) remains challenging. Because physical domains are naturally discretized as unstructured meshes rather than regular grids, standard convolutional architectures are often inadequate. In this paper, we propose a novel framework that extends diffusion posterior sampling (DPS) to graph-structured data. We develop an unconditional score-based diffusion model directly on a 2D triangular mesh to learn an accurate prior over the physical solution space. Furthermore, we introduce a regularized variant, RDPS, which incorporates explicit regularization terms, such as total variation and generalized Tikhonov, to complement the implicit diffusion prior and mitigate severe ill-posedness. Extensive experiments on synthetic and real 2D EIT datasets demonstrate that RDPS produces stable, physically plausible reconstructions. Our approach generalizes well to out-of-distribution inclusion geometries, is highly robust to measurement noise, and outperforms current state-of-the-art solvers (e.g., GPnP-BM3D, DP-SGS) in reconstruction accuracy and artifact reduction.
arXiv:2605.08391v2 Announce Type: replace
Abstract: Cooperative multi-agent reinforcement learning agents that act on partial local observations face a fundamental information bottleneck: the knowledge needed to select jointly optimal actions is scattered across the team, yet each agent must commit to a decision without access to its teammates' observations, intentions, or chosen actions. Existing methods either ignore this bottleneck, compress it into a scalar mixing signal, or route around it with learned communication channels. Framing action coordination as a problem of structured information integration among agents, we propose \textit{structured agent coordination via holistic information integration}, or SACHI, in which graph transformer convolutions over an inter-agent coordination graph enrich each agent's representation with receiver-sensitive, content-dependent signals from teammates prior to action selection. We evaluate SACHI across five cooperative tasks spanning spatial, communicative, and adversarial coordination challenges against twelve baselines. SACHI consistently matches or outperforms the best baseline on every task, and rigorous aggregate statistical analyses, including normalized metrics with bootstrap confidence intervals, Friedman ranking, and performance profiling, confirm that this advantage is statistically significant, robust across environments, and not attributable to increased model capacity. Parameter-matched ablations further trace the source of the gains to a single architectural property: the degree of content-dependence in the message-passing operator.
arXiv:2605.18813v1 Announce Type: new
Abstract: World models aim to predict plausible futures consistent with past observations, a capability central to planning and decision-making in reinforcement learning. Yet, existing architectures face a fundamental memory trade-off: transformers preserve local detail but are bottlenecked by quadratic attention, while recurrent and state-space models scale more efficiently but compress history at the cost of fidelity. To overcome this trade-off, we suggest decoupling future-past consistency from any single architecture and instead leveraging a set of specialized experts. We introduce a diffusion-based framework that integrates heterogeneous memory models through a contrastive product-of-experts formulation. Our approach instantiates three complementary roles: a short-term memory expert that captures fine local dynamics, a long-term memory expert that stores episodic history in external diffusion weights via lightweight test-time finetuning, and a spatial long-term memory expert that enforces geometric and spatial coherence. This compositional design avoids mode collapse and scales to long contexts without incurring a quadratic cost. Across simulated and real-world benchmarks, our method improves temporal consistency, recall of past observations, and navigation performance, establishing a novel paradigm for building and operating memory-augmented diffusion world models.
arXiv:2112.08507v5 Announce Type: replace
Abstract: Traditional randomized A/B experiments assign arms with uniform random (UR) probability, such as 50/50 assignment to two versions of a website to discover whether one version engages users more. To more quickly and automatically use data to benefit users, multi-armed bandit algorithms such as Thompson Sampling (TS) have been advocated. While TS is interpretable and incorporates the randomization key to statistical inference, it can cause biased estimates and increase false positives and false negatives in detecting differences in arm means. We introduce a more Statistically Sensitive algorithm, TS-PostDiff (Posterior Probability of Small Difference), that mixes TS with traditional UR by using an additional adaptive step, where the probability of using UR (vs TS) is proportional to the posterior probability that the difference in arms is small. This allows an experimenter to define what counts as a small difference, below which a traditional UR experiment can obtain informative data for statistical inference at low cost, and above which using more TS to maximize user benefits is key. We evaluate TS-PostDiff against UR, TS, and two other TS variants designed to improve statistical inference. We consider results for the common two-armed experiment across a range of settings inspired by real-world applications. Our results provide insight into when and why TS-PostDiff or alternative approaches provide better tradeoffs between benefiting users (reward) and statistical inference (false positive rate and power). TS-PostDiff's adaptivity helps efficiently reduce false positives and increase statistical power when differences are small, while increasing reward more when differences are large. The work highlights important considerations for future Statistically Sensitive algorithm development that balances reward and statistical analysis in adaptive experimentation.
arXiv:2512.01152v4 Announce Type: replace
Abstract: As we deploy machine learning systems in the real world, a core challenge is to maintain a model that is performant even as the data shifts. Such shifts can take many forms: new classes may emerge that were absent during training, a problem known as open-set recognition, and the distribution of known categories may change. Guarantees on open-set recognition are mostly derived under the assumption that the distribution of known classes, which we call the background distribution, is fixed. In this paper we develop CoLOR, a method that is guaranteed to solve open-set recognition even in the challenging case where the background distribution shifts. We prove that the method works under benign assumptions that the novel class is separable from the non-novel classes, and provide theoretical guarantees that it outperforms a representative baseline in a simplified overparameterized setting. We develop techniques to make CoLOR scalable and robust, and perform comprehensive empirical evaluations on image and text data. The results show that CoLOR significantly outperforms existing open-set recognition methods under background shift. Moreover, we provide new insights into how factors such as the size of the novel class influences performance, an aspect that has not been extensively explored in prior work.
arXiv:2605.19482v1 Announce Type: cross
Abstract: Radio technosignature searches and radio-based ultra-high energy (UHE) neutrino experiments address different scientific questions, but share a closely related data analysis problem: identifying rare signals of unknown morphology within large datasets dominated by thermal noise and anthropogenic radio-frequency interference (RFI). UHE neutrino radio experiments (including ARA, RNO-G, ANITA, and PUEO) have developed advanced methodologies for continuous-wave (CW) mitigation and background characterization. This invited contribution makes that connection concrete through three points. First, we demonstrate that catalog-based time-domain sine subtraction -- the CW mitigation technique used in ANITA and ARA -- can be adapted for technosignature pipelines by restricting subtraction to documented persistent contaminants, improving broadband transient visibility while preserving uncataloged narrowband candidates. Second, we identify a structural equivalence between spatiotemporal clustering used in UHE neutrino experiments and direction/cadence-based RFI rejection in radio SETI, proposing a joint feature space incorporating direction, time, frequency, bandwidth, duration, and polarization. Third, we argue that background-only anomaly ranking is the natural second stage of this workflow, providing morphology-agnostic candidate triage. Together, these ideas motivate a 'preserve-then-rank' workflow for commensal rare-event discovery, opening a near-term path toward cross-community collaboration.
arXiv:2605.18875v1 Announce Type: new
Abstract: In this short paper, we begin to investigate the conditions under which a generic Bipermutive Cellular Automaton (BCA) with no-boundary conditions of diameter $d$ generates a Latin square of order $N=2^{d-1}$ admitting an orthogonal mate, without relying on the linearity of the local rule. Since an orthogonal mate exists if and only if the Latin square can be partitioned into $N$ disjoint \emph{transversals}, we start by characterizing the subclass of BCA whose Latin squares have a transversal on the main diagonal. In particular, we prove that the main diagonal forms a transversal if and only if the generating function of the bipermutive local rule induces an invertible CA with periodic boundary conditions on a configuration of size $d-1$. We then perform exhaustive search experiments, showing that $d=6$ is the smallest diameter for which there exist nonlinear bipermutive CA that generate Latin squares with a transversal on the main diagonal.
arXiv:2603.01009v2 Announce Type: replace
Abstract: Over the past years, Automated Essay Scoring (AES) systems have gained increasing attention as scalable and consistent solutions for assessing the proficiency of student writing. Despite recent progress, support for Arabic AES remains limited due to linguistic complexity and scarcity of large publicly-available annotated datasets. In this work, we present Qayyem, a Web-based platform designed to support Arabic AES by providing an integrated workflow for assignment creation, batch essay upload, scoring configuration, and per-trait essay evaluation. Qayyem abstracts the technical complexity of interacting with scoring server APIs, allowing instructors to access advanced scoring services through a user-friendly interface. The platform deploys a number of state-of-the-art Arabic essay scoring models with different effectiveness and efficiency figures.
arXiv:2605.18805v1 Announce Type: new
Abstract: LLM recommendation agents increasingly produce structured recommendation reports: sets of items accompanied by natural-language justifications. Yet existing evaluations often reduce this setting to reranking small shortlisted candidate sets or judge reports mainly by semantic plausibility. We introduce Recommendation Atlas (Agentic Tool-Level Assessment for Shopping), or RecoAtlas, a benchmark and toolkit for evaluating shopping agents with behavior-grounded metrics. RecoAtlas complements held-out interaction metrics with learned utility proxies for relevance, complementarity, and diversity derived from interaction data, while separately measuring semantic coherence and explanation quality. Its controlled tool environment exposes agents to either semantic, behavior-aligned, or faulty tools, enabling diagnosis of whether performance gains arise from stronger reasoning, better signals, or more effective tool-use policies. Across controlled experiments, we show that RecoAtlas exhibits key properties of a meaningful benchmark for agentic systems: performance scales with model capacity and test-time compute, improves with stronger and better-aligned tools, degrades under noisy or misaligned signals, and reveals that semantic plausibility does not necessarily capture behavior-grounded utility. RecoAtlas provides a foundation for developing and evaluating shopping assistants that optimize not only for plausible recommendations, but also for coherent, behaviorally grounded recommendation sets.
arXiv:2605.19008v1 Announce Type: new
Abstract: Modern language-model training is increasingly exposed to instability, degraded runs, and wasted compute, especially under aggressive learning-rate, scale, and runtime-stress conditions. This paper introduces Learn-by-Wire Guard (LBW-Guard), a bounded autonomous training-control governance layer that operates above AdamW. Rather than replacing the optimizer update rule, LBW-Guard observes training telemetry, interprets instability-sensitive regimes, and applies bounded control to optimizer execution while preserving fixed training objectives.
We evaluate LBW-Guard in a Qwen2.5-centered stress-and-robustness suite using WikiText-103, with Qwen2.5-7B as the empirical anchor, model-size comparisons against Qwen2.5-3B and Qwen2.5-14B, learning-rate stress tests, gradient-clipping baselines, and a no-LoRA TinyLlama-1B full-parameter sanity check. In the 7B reference setting, LBW-Guard reduces final perplexity from 13.21 to 10.74, an 18.7% improvement, while reducing end-to-end time from 392.54s to 357.02s, a 1.10x speedup. Under stronger learning-rate stress, AdamW degrades to 1885.24 final perplexity at LR=3e-3 and 659.76 at LR=1e-3, whereas LBW-Guard remains trainable at 11.57 and 10.33, respectively. Gradient-clipping baselines do not reproduce this effect.
These results support a scoped systems conclusion that stability-sensitive LLM training can benefit from a governance plane above the optimizer. LBW-Guard provides evidence that bounded runtime control can preserve productive compute under stress while remaining distinct from optimizer replacement and local gradient suppression.
arXiv:2605.19556v1 Announce Type: new
Abstract: Estimating relative pose from image pairs fundamentally requires only a minimal subset of geometrically consistent correspondences. However, most learning-based approaches rely on dense matching or direct regression, leading to redundancy and reduced geometric interpretability. In this work, we propose a sparse epipolar matching framework that predicts a compact set of correspondences optimized for geometric consistency across varying temporal baselines. To address residual noise and misalignment, we introduce an epipolar diffusion process that models correspondence uncertainty and refines keypoints toward epipolar consistency. The refined correspondences, along with depth cues, are lifted into a graph representation forming a Steiner graph that encodes relational structure between points. A graph neural network learns a compact subset of informative correspondences, which are passed to a differentiable singular value decomposition solver for end-to-end geometric estimation. Relative pose is recovered from the resulting essential matrix and evaluated in a visual odometry setting on the TartanAir and KITTI SLAM datasets. Experimental results demonstrate that combining sparse matching, diffusion-based refinement, and graph-based subset selection reduces correspondence redundancy while maintaining robust pose estimation across challenging baselines.
arXiv:2409.03192v2 Announce Type: replace
Abstract: Fine-grained image classification has witnessed significant advancements with the advent of deep learning and computer vision technologies. However, the scarcity of detailed annotations remains a major challenge, especially in scenarios where obtaining high-quality labeled data is costly or time-consuming. To address this limitation, we introduce Precision-Enhanced Pseudo-Labeling(PEPL) approach specifically designed for fine-grained image classification within a semi-supervised learning framework. Our method leverages the abundance of unlabeled data by generating high-quality pseudo-labels that are progressively refined through two key phases: initial pseudo-label generation and semantic-mixed pseudo-label generation. These phases utilize Class Activation Maps (CAMs) to accurately estimate the semantic content and generate refined labels that capture the essential details necessary for fine-grained classification. By focusing on semantic-level information, our approach effectively addresses the limitations of standard data augmentation and image-mixing techniques in preserving critical fine-grained features. We achieve state-of-the-art performance on benchmark datasets, demonstrating significant improvements over existing semi-supervised strategies, with notable boosts in accuracy and robustness.
arXiv:2605.18812v1 Announce Type: new
Abstract: Modern NLP and LLM systems are pipelines: named entity recognition (NER) -> entity disambiguation (NED) -> entity typing, retrieval-augmented generation (retriever -> reader), and agentic chains of planner -> tool -> critic. Errors compound across stages, but existing uncertainty quantification methods either calibrate each stage independently (no joint coverage) or apply a Bonferroni union bound (joint coverage, but conservative). We present PASC (Pipeline-Aware Split Conformal), which reduces multi-stage joint coverage to a single scalar conformal prediction problem on the joint maximum nonconformity score. PASC provides a finite-sample distribution-free guarantee that all K stages are simultaneously covered with probability at least 1 - alpha, and is nearly tight up to a 1/(n+1) factor. On a three-stage NER -> NED -> entity-typing pipeline over CoNLL-2003, PASC achieves 96.4% end-to-end coverage versus 93.4% for Bonferroni and 86.5% for independent CP, at identical average prediction set size (1.083). Under distribution shift to WNUT-17 Twitter and WikiNEuRal Wikipedia data, PASC empirically maintains the target coverage in the tested shift settings while independent CP collapses to 59%. PASC requires a single quantile computation, runs 1.7x faster than Bonferroni, and scales to K = 6 stages where independent CP drops to 0.53 end-to-end coverage. The same joint-maximum-score reduction applies directly to compound LLM systems and agent pipelines.
arXiv:2605.19564v1 Announce Type: new
Abstract: The rise of Physics-Informed Neural Networks (PINNs) has popularized the concept of solving differential equations via residual minimization. However, neural networks are often viewed as ``black boxes" requiring significant computational overhead and stochastic optimization. Moreover, PINNs typically treat boundary conditions (BCs) as ``soft constraints" within the loss function and this makes the optimization process struggling to enforce the BCs properly. This paper introduces the \textbf{Spline-based Physics-Informed Numerical Scheme (SPINS)}, a numerical framework designed to solve both initial and boundary value problems of ordinary differential equations (ODEs). By replacing the neural network architecture of traditional PINNs with a structured spline basis, SPINS achieves high accuracy and interpretability with a minimal parameter set. In addition, the BCs are automatically satisfied from the choice of the splines architecture. Therefore, SPINS provides smooth numerical solutions for ODEs allowing analytical differentiation. Moreover, SPINS benefits from the automatic differentiation where computing the gradient of the physics-informed loss function is an easy task making the optimization process very fast using gradient-based optimizers such as the L-BFGS-B algorithm. We demonstrate the efficacy of SPINS on nonlinear second order ODEs with several choices of BCs using cubic and quintic interpolating splines and present its natural extension to high order ODEs.
arXiv:2408.06843v5 Announce Type: replace
Abstract: We present an efficient task and motion replanning approach for sequential multi-object manipulation in dynamic environments. Conventional Task And Motion Planning (TAMP) solvers experience an exponential increase in planning time as the planning horizon and number of objects grow, limiting their applicability in real-world scenarios. To address this, we propose learning problem decompositions from demonstrations to accelerate TAMP solvers. Our approach consists of three key components: goal decomposition learning, computational distance learning, and object reduction. Goal decomposition identifies the necessary sequences of states that the system must pass through before reaching the final goal, treating them as subgoal sequences. Computational distance learning predicts the computational complexity between two states, enabling the system to identify the temporally closest subgoal from a disturbed state. Object reduction minimizes the set of active objects considered during replanning, further improving efficiency. We evaluate our approach on three benchmarks, demonstrating its effectiveness in improving replanning efficiency for sequential multi-object manipulation tasks in dynamic environments.
arXiv:2408.01896v5 Announce Type: replace
Abstract: The idea of security sharing goes back to Nakamoto's introduction of merge mining, a technique that enables Bitcoin miners to reuse their hash power to bootstrap and secure other Proof-of-Work (PoW) blockchains. However, with the rise of Proof-of-Stake (PoS) chains, there is a need for new methods of Bitcoin security sharing. We introduce Bitcoin staking, a protocol that allows Bitcoin holders to trustlessly use their idle asset to secure a PoS chain. The key challenge is to enable automatic slashing of bitcoins on the Bitcoin chain upon safety violations on the PoS chain. We achieve this using double-authentication-preventing signatures, finality gadgets and bi-directional timestamping between Bitcoin and the PoS chain. Our design is entirely modular and can be integrated with any PoS chain. A version of this protocol was deployed to secure the Babylon mainnet in April 2025 and currently has over 58,000 bitcoins staked (about 4 billion USD at current prices) while paying only 0.05% APR reward to the stakers. This is 2 orders of magnitude cheaper security cost than in PoS chains secured by their native token.
arXiv:2510.03485v2 Announce Type: replace
Abstract: Autonomous web agents are increasingly deployed for long-horizon tasks, yet their ability to adhere to real-world policies remains critically underexplored compared to standard safety objectives. To address this gap, we introduce PolicyGuardBench, a benchmark of 60k policy-trajectory pairs designed to evaluate compliance through both full-trajectory and novel prefix-based violation detection tasks. Using this dataset, we train PolicyGuard, a lightweight guardrail model that achieves strong detection accuracy while maintaining high inference efficiency. Notably, our model demonstrates robust generalization capabilities, preserving high performance even on unseen domains. These contributions establish a comprehensive framework for studying policy compliance, showing that accurate and generalizable guardrails are feasible at small scales.