arXiv:2607.01106v2 Announce Type: replace
Abstract: Distributed trajectory estimation arises in many applications across robotics, but existing implementations typically do not consider asynchrony in agents' communications and computations. Therefore, we propose an asynchronous block coordinate descent algorithm for distributed trajectory estimation. We consider a team of agents that observes a team of robots and estimates the robots' states over a sliding window. The agents solve an approximation of the maximum a posteriori estimation problem, which we derive. We show this approximation introduces negligible errors and eliminates up to 96.9% of communications among agents. Next, we prove that agents' iterates converge exponentially fast to the optimal estimate of the robots' states. Simulations show that this approach has up to 64% less error than a comparable state-of-the-art algorithm. Experiments on mobile robots show this approach is robust to delays whose lengths span three orders of magnitude.
Science Journals
arXiv:2607.01153v2 Announce Type: replace
Abstract: Safety evaluations for language models increasingly depend on judgments about ambiguous natural-language behaviour: whether a model has followed an instruction, refused appropriately, complied with a policy, resisted an embedded command, or misreported progress in an agentic task. Existing benchmarks often compress these distinctions into pass/fail labels, obscuring whether failures arise from capability limits, policy ambiguity, instruction conflict, scaffold failure, or unstable evaluator judgments.
This paper introduces adversarial pragmatics as a benchmark and annotation protocol for evaluating model behaviour under instruction conflict, embedded commands, quotation, scope ambiguity, deixis, indirect speech acts, and multi-turn agent transcripts.
The contribution is empirical and methodological: a linguistically controlled taxonomy, an 18-item seed benchmark with validator-enforced metadata, a 54-row local seed pilot, an expert-evaluation protocol distinguishing task success, policy compliance, safety risk, refusal outcome, and evaluator confidence, and metrics for judge validity, diagnostic ambiguity, and taxonomy drift. The benchmark treats labels as inference licenses: it tests whether safety-relevant categories project across paraphrase, wrapper, model, and judge condition. In the pilot, a rubric-aided LLM judge graded its own outputs with expected-behaviour fields visible and still missed the safety-relevant minority classes.
arXiv:2607.01211v2 Announce Type: replace
Abstract: Repository-level performance-optimization benchmarks such as GSO, SWE-Perf and SWE-fficiency evaluate coding agents by applying patches to real repositories and comparing runtime against unoptimized baselines and official reference patches. Their leaderboard scores are increasingly used as evidence of coding-agent progress, but those scores can conflate runtime instability, benchmark-specific scoring rules, and how many tasks are already solved by at least one public submission. We audit these issues across the three benchmarks. First, we replay the official reference patches for 740 code optimization tasks across four common types of Google Cloud machines. Most benchmark tasks can be replayed, but their reference patches satisfy the original benchmark validity rules in every cross-machine replay for only 39/102 GSO tasks, 11/140 SWE-Perf tasks, and 411/498 SWE-fficiency tasks; SWE-Perf is especially fragile because many reference patches produce close-to-zero runtime changes. Second, we show that public submission rankings depend strongly on the benchmark scoring rule. Among eight public submissions shared by GSO and SWE-fficiency, the official rankings disagree on 9 of 28 pairwise submission comparisons, and SWE-fficiency's leaderboard scoring rule assigns the worst ten tasks overly high score weights of 58.5%-82.8%. Third, looking across 10 public submissions for each task, we find that at least one submission matches or beats the reference patch on 85.3% (384/450) of replay-valid GSO and SWE-fficiency tasks, and beats the unoptimized base code on 99.8% (449/450). Our study complements leaderboard scores by identifying tasks with more reliable performance signals, quantifying per-task score contributions, and exposing the remaining performance gaps that are hidden by aggregate rankings.
arXiv:2607.01578v2 Announce Type: replace
Abstract: 3D Gaussian Splatting (3DGS) enables real-time novel view synthesis for static scenes. Extending it to dynamic scenes via deformation fields has recently attracted significant attention, particularly for dynamic scene reconstructionband distractor-free. However, existing deformation networks lack explicit motion awareness: they neither capture long-term motion intensity nor exploit short-term temporal coherence, leading to inaccurate foreground deformation and pseudo-static residuals in the background. We present MVFusion-GS, a method that enhances deformation networks with two complementary motion-aware mechanisms. The Motion-Variance Guided Refinement aggregates per-Gaussian deformation statistics across time to estimate motion variance and uses it to guide dynamic-static separation during deformation prediction. The MotionFormer Temporal Attention module applies Transformer self-attention over neighboring timesteps to model local motion dependencies and improve temporal consistency. Extensive experiments on both dynamic scene reconstruction and distractor-free reconstruction benchmarks demonstrate state-of-the-art performance, showing that explicit motion awareness improves both foreground motion modeling and static background reconstruction.
arXiv:2607.03454v2 Announce Type: replace
Abstract: In this paper, we propose Adversarial Dynamics Priors (ADP) for perturbation-resilient humanoid locomotion control. Existing motion prior-based methods induce natural motion styles by imitating kinematic motion features, but they do not directly regularize dynamics features, such as CoM motion, centroidal momentum, contact forces, and contact states. To address this limitation, we replace kinematic motion-style feature with selected dynamics features extracted from locomotion trajectories as the target of adversarial regularization. To this end, we use trajectory optimization to construct a reference dataset and train a discriminator to evaluate whether policy-induced temporal windows are consistent with the resulting reference distribution. Without explicit motion tracking, ADP encourages policy rollouts to remain close to the reference support, even after perturbations. Experimental results show that, compared with AMP, the strongest baseline in our evaluation, ADP improves the $80\%$-success impulse threshold ($J_{80}$) by $16.7\%$, while reducing direction-averaged recovery time and velocity tracking error by $47.9\%$ and $35.4\%$, respectively.
arXiv:2607.04667v2 Announce Type: replace
Abstract: Software engineering has a complicated relationship with "correctness". We recognize the challenges of full formal rigor as well as many required properties beyond functional correctness. Although we satisfice in practice, we are still stuck in the mindset that we could reason our way to correctness, if only we had enough information. Unfortunately for our hopes of formal rigor, our expectations are shaped by unspoken knowledge that is personal, subjective, qualitative, and largely unavailable. Generative AI has introduced a new dimension to assurance: its foundation is statistical rather than formal. Traditional software engineering establishes confidence through rigorous reasoning, domain knowledge and expert judgment. In contrast, generative AI results are sophisticated predictions, "probably approximately correct". This inherently limits assurances about the results to probabilistic assertions. Further, the nuances that guide human judgment are often tacit or implicit. This knowledge casts only scant shadows into the digital record, so that critical source of knowledge is only faintly represented in AI models. We have many approaches for developing assurances that a software system does what it's expected to do, though most of them focus on code specifications rather than system requirements, let alone the system's fitness for its purpose. We have failed to develop a systematic understanding of the relative merits of the various approaches to assurance. I hope that generative AI will finally force us to tackle this. To that end, I will challenge us to think systematically about our assurance techniques, especially the role of hidden context and the challenges of AI. We need ways to make informed, reasoned choices about cost-effective combinations of approaches to developing confidence in our systems. We call ourselves software engineers. Let's act like engineers.
arXiv:2607.04763v2 Announce Type: replace
Abstract: We study on-policy distillation (OPD) for agentic tasks, where an LLM agent interacts with an environment over multiple turns and a student imitates a teacher over these multi-turn interaction histories. Fully online OPD is costly because each update requires fresh student rollouts through the environment and teacher queries at visited histories. We propose Replayed-Prefix On-Policy Distillation (ReOPD), an off-environment alternative that reuses pre-collected teacher trajectories as replayed prefixes: the student acts at selected steps, while the teacher provides dense per-step supervision without executing new environment interactions. We show that multi-turn OPD introduces a prefix trap: making histories more student-on-policy improves relevance to the student, but can query the teacher on histories where its target is unreliable. This creates a two-sided distribution shift between student occupancy and teacher reliability. ReOPD addresses this by treating multi-turn OPD as a reliability-aware prefix distribution design and implements it with a simple step-decaying sampling schedule that emphasizes early, lower-shift prefixes. Across mathematical reasoning with Python and search environments over multiple teacher and student model scales, ReOPD preserves or improves OPD-level accuracy, uses zero tool calls during student training, and is at least 4$\times$ faster per rollout than OPD. ReOPD therefore turns expensive agent-environment interaction into a reusable offline resource, enabling scalable distillation across tools, tasks, and environments.
arXiv:2607.06214v2 Announce Type: replace
Abstract: This paper offers a framework for considering curiosity as an ecosystem. First, it suggests that a single agent's inquiry policy (how, when, and why an agent asks a question) depends on how the agent values immediate uncertainty reduction, costs, delayed return, and the value of keeping the question open. A key concept in the framework is that the weights on these decision-related terms can change with experience. For example, a period of cheap, quickly answered questions may change the cost of inquiry on a short timescale and change which kinds of questions the agent is drawn to answer over a longer timescale. Second, these ideas are extended to many agents exploring a shared knowledge landscape, and there the framework tracks inquiry volume, topic diversity, frontier-directed inquiry, redundancy, and reusable knowledge. The result is a conceptual framework for studying curiosity ecology and for future efforts towards designing multi-agent AI systems for discovery.
arXiv:2607.06864v2 Announce Type: replace
Abstract: We give randomized $(5+\epsilon)$-approximation algorithms for both the continuous and discrete Fr\'echet distances on arbitrary two polygonal curves $\tau$ and $\sigma$ in $\mathbb R^d$ for fixed $d$, with $n$ and $m\le n$ vertices respectively. Our algorithm for continuous Fr\'echet runs in $\widetilde O_{d,\epsilon}(n m^{8/9})$ time, and our algorithm for discrete Fr\'echet runs in $\widetilde O_{d,\epsilon}(n m^{4/5})$ time. These bounds improve the recent strongly subquadratic constant-factor approximation algorithms of Cheng, Huang, and Zhang~\cite{cheng2025constant}, which give $(7+\epsilon)$-approximations.
The approximation improvement comes from certifying long boundary-to-boundary reachability directly through auxiliary surrogate curves, avoiding an extra conversion back to input subcurves and hence removing one triangle-inequality loss. The running-time improvement comes from a two-scale macro-surrogate search combined with dyadic auxiliary-transfer structures, with the discrete case gaining a faster bound from exact planar reachability in the discrete free-space graph.
arXiv:2607.08208v2 Announce Type: replace
Abstract: This paper describes our self-designed system for Task 1 of the MLC-SLM 2026 Challenge for multilingual two-speaker conversational speech. The system combines a modular speaker diarization front end with a challenge-adapted Qwen3-ASR-1.7B recognizer. The diarization front end performs voice activity detection, subsegment generation, CAMPPlus speaker embedding extraction, two-speaker spectral clustering, and RTTM-based audio segmentation. The resulting speaker-attributed segments are grouped by language or region and decoded by the adapted ASR model. For ASR adaptation, we first perform supervised full fine-tuning on the official training data, then apply LoRA fine-tuning with synthetic speech generated by a three-pipeline TTS-based synthetic speech augmentation framework, and finally refine the model using GRPO reinforcement learning with rewards based on WER/CER and penalties for hallucination, repetition, and length deviation. On the official development set, the full system achieves an average tcpMER of 23.70, reducing the error rate by 6.83 absolute points relative to the released Qwen-ASR-1.7B performance. On the final evaluation set, the system achieves an average tcpMER of 17.97. Ablation results show that supervised fine-tuning provides the largest gain, while synthetic-speech LoRA adaptation and reinforcement learning further improve robustness.
arXiv:2607.08748v2 Announce Type: replace
Abstract: In this study, we present a large-scale descriptive analysis of the use of an AI-based learning assistant (Syntea) in higher education. Based on objective log data from 77,543 students enrolled in distance studies, we examine usage patterns across gender, age group, study cluster, degree, and study mode. To date, existing research on educational chatbots has largely relied on comparatively small samples and self-reported survey data, while large-scale evidence on actual usage behavior remains limited. Our findings show that Syntea is already embedded in the study routines of many learners, but that usage differs across demographic and structural contexts. By identifying these patterns, our study provides an empirical basis for the further development of AI-based learning support and contributes a large-scale analysis of educational chatbot usage in higher education.
arXiv:2607.09142v3 Announce Type: replace
Abstract: Large language models (LLMs) are increasingly deployed in online medical consultation, yet existing benchmarks remain poorly aligned with real clinical practice. Many rely on synthetic conversations or patient simulators, omit patient-uploaded medical images, or evaluate open-ended clinical responses using multiple-choice or lexical-overlap metrics that poorly reflect clinical quality. We introduce \textbf{MedRealMM}, a large-scale benchmark for multimodal online medical consultation built from de-identified patient-doctor interactions collected from a nationwide Chinese internet hospital. MedRealMM uses a Multimodal Clinical Challenge Point (MCCP) extraction framework to identify clinically demanding moments in authentic consultation trajectories and converts each into a standardized next-response generation task while preserving the preceding text-image context. Each instance is paired with a case-specific rubric refined by physicians that rewards clinically desirable behaviors and penalizes unsafe, unsupported, or contradictory responses. The current release contains 5,620 real-world multimodal cases spanning 64 clinical departments. We evaluate 19 general-purpose and medical-specialized LLMs, including text-only and multimodal systems. Our results show that image information is critical for reliable clinical performance and that current frontier models remain below the online physician response. Although some frontier models satisfy as many or more positive clinical criteria than physicians, they trigger more negative criteria, indicating that safety-sensitive error avoidance remains a central bottleneck. MedRealMM offers a realistic and reproducible benchmark for evaluating multimodal medical reasoning in real-world online consultation. The dataset will be publicly available on Hugging Face at https://huggingface.co/datasets/jdh-algo/MedRealMM.
arXiv:2607.09449v2 Announce Type: replace
Abstract: Bayesian causal discovery is widely used for its ability to quantify epistemic uncertainty over directed acyclic graphs (DAGs) through posterior inference. However, its behaviour under latent confounding remains poorly understood, as existing work typically notes that confounding breaks identifiability without characterising how the posterior distribution over DAGs responds. In this work, we analyse posterior behaviour under latent confounding in linear Gaussian causal models, focusing on additive latent confounding between exactly two observed variables. We derive a critical correlation threshold above which the score function favours graphs with a spurious edge between the confounded variables, and show that this threshold decreases with sample size -- more data lowers the correlation required for the spurious edge to be favoured. Beyond this threshold, we characterize two distinct posterior failure regimes determined by the local structure around the confounded variables. Our findings are supported by exact posterior computations on multiple graph structures, demonstrating both the predicted failure regimes.
arXiv:2607.09540v3 Announce Type: replace
Abstract: The evolution of recommender systems can be explored by asking how they utilize information at scale. Throughout most of the historical period under consideration during the past two decades, industrial systems have relied on raw IDs, which are discrete, globally unique, and semantically opaque identifiers that enable exact lookup, logging, and item-specific memorization at scale. Over time, however, recommender systems have sought to utilize richer sources of information, including item content, context, multimodal signals, and cross-domain structure. This development has led to a new stage in which part of such information is no longer used solely as auxiliary features around item identity, but is increasingly encapsulated in semantic IDs that provide a more structured, model-facing form of identity. We argue that this shift goes beyond the rise of generative recommendation over traditional methods. Indeed, it reflects a broader evolution in how recommender systems utilize information under industrial-scale constraints. This paper looks at the past, present, and future to examine three connected questions: why raw IDs dominated the early development of recommender systems, why semantic information is increasingly being encapsulated in IDs today, and what may come next once recommendations move beyond semantic retrieval. In particular, we introduce semantic planning as a possible future direction in which the system first predicts the semantic target of the next exposure, and only then instantiates that target as a specific item or generated creative. We further argue that such a shift may require changes not only in model design but also in evaluation and in the way recommender systems coordinate the objectives of users, platforms, and providers.
arXiv:2607.09815v2 Announce Type: replace
Abstract: A central challenge in image-goal UAV navigation under Global Navigation Satellite System (GNSS) denial is estimating metric distance and heading between current and goal views. Dense pairwise geometry models capture relative scene structure, but without a calibrated metric scale, they cannot directly provide reliable distance estimates for navigation. Although global scale calibration corrects the dominant scale bias, the remaining errors vary systematically with distance. In this paper, Range-Aware Scale Recovery (RASR) is proposed, which complements global scale calibration with range-aware residual correction. RASR encodes pairwise geometry extracted by a frozen Matching And Stereo 3D Reconstruction (MASt3R) backbone as a compact descriptor and separates the scale-recovery core from task-specific command calibration. On the official online evaluation of the UAVs in Multimedia 2026 PairUAV challenge, RASR achieved a total error of 0.003189, achieving a lower total error than global scale calibration alone. The results demonstrate that range-aware residual correction improves metric distance estimation beyond global scale calibration. Code and materials are available at https://github.com/lht-research/rasr-pairuav.
arXiv:2607.09957v2 Announce Type: replace
Abstract: This report studies on-device English-to-Traditional-Chinese subtitle translation for Taiwan under short inputs, short outputs, batch-size-one inference, low latency, and privacy constraints. These conditions limit the value of optimizations designed for long-context or high-throughput language-model serving. Starting from LMT-60-0.6B, preliminary profiling suggests that vocabulary projection becomes a more important decode-time cost after GGUF quantization reduces the relative cost of Transformer blocks. We replace the original 151k-token vocabulary with a 64k-token subtitle-domain tokenizer, migrate the embedding space, and adapt the model through embedding calibration followed by full supervised fine-tuning. On an OpenSubtitles2024 test set, LocalSubs achieves a 59.2% tie-excluded win rate against Google Translate under GPT-4o pairwise judging. Performance is strongest on short cues and declines as cue length increases. In a separate preliminary Apple M2 Metal profiling run, LocalSubs shows a 1.63x speedup over a 151k-vocabulary baseline. The code is available on https://github.com/aiden1020/localsubs .
arXiv:2607.10079v3 Announce Type: replace
Abstract: Digital Adoption Platforms (DAPs) are embedded overlays widely used on web systems to guide users through operations inside a page, helping them get started with unfamiliar interfaces quickly. Completing a real task, however, rarely means clicking a few buttons on a single page: it takes a sequence of actions that unfolds across changing page states. Prior studies have also treated automated web agent actions and guide text generation as two separate problems, and most of them feed models textual page representations such as the DOM or accessibility trees rather than the rendered screens that humans actually operate on. In this work we introduce MAG, the first benchmark that unifies task execution and guide writing into a single Multimodal Action and Guide task, with two grounding schemes over screenshots: Set-of-Mark element selection and raw pixel coordinates. We further build a complete harness for this compound task, covering annotation with LLM assistance and human verification, training, evaluation in live environments, and joint metrics for actions and guides. With this harness we evaluate frontier API models and open multimodal models, and report detailed analyses. Finally, we design a GRPO training method augmented with expert trajectories, which nearly doubles the success rate of a supervised 9B agent (from 6.9% to 13.2%) and improves guide quality at the same time. Even the strongest model completes fewer than 40% of the tasks, leaving ample room for future research.
arXiv:2607.10132v2 Announce Type: replace
Abstract: Dynamic loco-manipulation requires legged robots to coordinate whole-body motion while maintaining stable physical interaction with grasped objects under uncertain external forces. While tactile sensing has been widely studied for robotic manipulation, its role in dynamic whole-body control remains largely unexplored. Existing works without tactile feedback commonly grasp firmly rather than regulate the grasp according to the interaction. We propose TAC-LOCO, a tactile-augmented unified reinforcement learning framework that encodes tactile array observations from compliant grippers into a compact latent representation and joins it with proprioception for unified control of the legs, arm, and gripper. With effective grasp stability reward design, the policy learns to simultaneously track body velocity and end-effector trajectories, moderate grasp force, and prevent object slip under both gradual load changes and sudden release events. We deploy the policy zero-shot on a Unitree Go2 with an Interbotix WidowX 250 arm and tactile gripper, demonstrating dynamic tactile-informed loco-manipulation under varying external interactions, achieving a 47% reduction in grasping force and an object drop rate of less than 1%.
arXiv:2607.10255v2 Announce Type: replace
Abstract: Min-wise hashing and its $k$-min-wise variant are standard tools in similarity estimation, sampling, sketching, and streaming. A $k$-min-wise family requires every prescribed $r$-subset of a fixed set, for $r\le k$, to appear as the $r$ smallest hash values with approximately the fully random probability, up to multiplicative error $\delta$. Previous analyses show that $O(\log(1/\delta)+k\log\log(1/\delta))$-wise independence suffices. Consequently, for $k=\Theta(\log N)$ and $\delta=N^{-c}$, the standard polynomial construction uses $O(k\log N\log\log N)$ seed bits. Recent work of Chen, Huang, and Li achieves the optimal $O(k\log N)$ seed length for $k=\log^{O(1)}N$, but only with almost-polynomial error $2^{-O(\log N/\log\log N)}$, leaving open whether polynomially small error is possible with the same seed length. We prove that the standard $s$-wise independent polynomial hash family is $k$-min-wise with multiplicative error $\delta$ for $s=O(k+\log(1/\delta)).$ Thus, when $k=\Omega(\log(1/\delta))$, only $O(k)$-wise independence is required. In particular, for $k=\Theta(\log N)$ and $\delta=N^{-c}$, this gives an explicit family with seed length $O(k\log N)$, matching the support-size lower bound up to constant factors. The proof conditions on the prescribed bottom set and bounds the error only after averaging over the random threshold given by its largest hash value, rather than controlling every threshold separately.
arXiv:2607.10310v2 Announce Type: replace
Abstract: Preparing for job interviews is important for securing desired positions, yet realistic practice remains difficult to access: real interviews are infrequent, expert mock coaching is costly, and self-practice offers neither adaptive dialogue nor structured assessment. Existing systems typically address only parts of this need through fixed question sequences, limited communication channels, or feedback with little supporting evidence. We present PolyInterview, an LLM-based platform for immersive mock interview practice with comprehensive multimodal assessment. PolyInterview uses the target job description and CV to generate questions tailored to the role and candidate, conducts multi-turn spoken interviews with a lip-synced digital human interviewer that asks answer-aware follow-up questions, and evaluates response content, vocal delivery, and non-verbal behavior. Four parallel evaluators produce 13 behavior-level features that are aggregated into 10 assessment aspects and two competency tracks. Guided by the KSA and STAR frameworks, the report links each score to behavioral evidence and actionable recommendations. PolyInterview is publicly accessible. Its current all-account snapshot contains 101 accounts, 1,564 interview sessions, 7,665 generated questions, and 1,422 five-stage question sets. Generated questions are more closely aligned with their matched job description than with cross-role job descriptions in 93.7% of sessions. An evaluation by ten experts found strong question plans and actionable feedback.
arXiv:2607.10847v2 Announce Type: replace
Abstract: Ehrenfest dynamics is a widely used mixed quantum--classical approach for nonadiabatic molecular dynamics, whereas thawed Gaussian wavepacket dynamics provides an efficient semiclassical description of adiabatic nuclear quantum dynamics. Here we describe thawed Gaussian Ehrenfest dynamics (TGED), which unifies and generalizes these two methods to capture both electronic nonadiabaticity and nuclear quantum effects within a single framework. The fully variational formulation of TGED is derived by applying the time-dependent variational principle to a Hartree product of electronic and Gaussian nuclear wavepackets. Replacing the effective locally quadratic molecular potential obtained from this variational treatment by alternative effective locally quadratic potentials yields an infinite family of TGED methods, of which we present several members. We analyze the limiting cases of the general formalism and show, in particular, that it reduces to conventional Ehrenfest dynamics in the classical limit for the nuclei and to thawed Gaussian wavepacket dynamics in the absence of electronic coupling. Finally, we present explicit geometric integrators for the entire family of methods and identify the conditions under which the different approximations become exact.
arXiv:2607.11149v2 Announce Type: replace
Abstract: LLM agent benchmarks measure task completion, reliability, and inference cost, but not the persistent data an agent run leaves on disk, including logs, context snapshots, checkpoints, and debug traces. We introduce AgentFootprint, a cross-framework benchmark of post-run agent storage footprint. Its serialization-aware metric suite measures total retention, channel composition, duplication, growth, compressibility, and conversation-history reconstructability. It addresses a measurement trap: naive byte-level measurement understates duplication by an order of magnitude because database paging and JSON escaping obscure repeated content. A fixed-trace control separates agent-generated logical volume from persistence-layer amplification: replaying the same trajectory through seven persisting frameworks yields a 6.7x spread. Under identical models, tools, and tasks, configurations with 100% accuracy differ by 15.7x in retained bytes, although their defaults support different recovery and audit capabilities. Three full-history configurations grow superlinearly on a repeated-observation stress task. Exported trajectories from 108 instance-normalized SWE-bench Verified submissions span three orders of magnitude per instance, with no detectable correlation with resolve rate. A content-addressed store reduces retention by 4.8x-32.7x while preserving every reconstructability score. These results establish persistent storage as a resource metric to report jointly with accuracy and reconstructability.
Compile, Then Page: Executable SOP Programs and a Capability-Gated Runtime for Procedural LLM Agents
arXiv:2607.11346v2 Announce Type: replace
Abstract: Enterprise agents must follow long-horizon, conditional, safety-critical standard operating procedures (SOPs). We compile machine-readable SOP constraints into executable pseudo-code and run them with a program-guided (PG) stack machine that pages the active frame while an LLM performs semantic execution. A three-arm SOPBench study across six models separates representation from runtime: compiled text never significantly hurts and gains up to 16.0 points where official prose underperforms. Runtime guidance is capability-gated. Two strong models independently show positive seven-domain PG contrasts (58:19 and 75:31 discordant pairs), whereas weak models are harmed. A full-program cursor ablation (active frame first, complete program retained) recovers much of the strong-model refusal gain; selective visibility adds a smaller improvement. Paired probe and audit measurements track this divide to spontaneous state discipline rather than reconstruction ability. On Bank the three primary arms rise from 70.4 to 86.4 to 92.8, with 100% refusal correctness. Practical guidance: compile first; enable active-frame paging only after a model-level discipline check.
arXiv:2607.11623v2 Announce Type: replace
Abstract: Real-world paired image dehazing remains challenging because haze degradation is spatially non-uniform, illumination-dependent, and physically ambiguous even when haze-free references are available. Existing end-to-end restoration networks usually learn a deterministic mapping from a hazy observation to a clean target, while degradation-sensitive feature responses, reverse haze-formation consistency, and cross-domain negative structure remain insufficiently exploited. In this paper, we propose Backbone-Agnostic Stochastic Perturbation Learning (BSPL), a plug-and-play framework for end-to-end real-world image dehazing. BSPL first introduces a Learnable Stochastic Perturbation Modulator (LSPM), which learns input-conditioned channel-wise and spatial-wise perturbation distributions and converts the resulting feature-response discrepancies into adaptive modulation weights. It then develops a Prior-informed Perturbation-guided Reconstruction Module (PPRM), which reuses the learned bottleneck perturbations together with transmission and atmospheric-light priors to reconstruct the hazy observation from the restored result and enforce degradation consistency. Furthermore, we propose a Dual-space Domain-diversified Distribution-aware Contrastive Loss ($D^3$CL) to regularize both clean restoration and hazy reconstruction spaces with real-world and synthetic negatives. Experiments on five real-world paired benchmarks show that BSPL consistently improves multiple representative backbones with only marginal additional inference overhead.
arXiv:2607.11624v3 Announce Type: replace
Abstract: Reinforcement learning (RL) algorithms classically suffer from poor sample efficiency. In robotics, a recent line of work has emerged addressing this problem by encoding physics priors in the learning process. However, most of these approaches are validated on well-defined, low-dimensional benchmark systems rather than high-dimensional robots with complex nonlinear dynamics. In this paper, we introduce \textit{SKooP (Symmetric Koopman Predictions)}, an approach combining the advantages of morphological symmetries with those of a Koopman model learned via autoencoder to enhance policy learning. SKooP learns a Koopman model of the system dynamics alongside the policy. The resulting Koopman predictions are used as privileged observations for the critic, allowing the agent to learn based on smoother, more informative features. We also incorporate group symmetries into the actor, critic, encoder and decoder networks to produce a highly equivariant policy. The SKooP approach is validated via in-depth analysis of the learned Koopman models and symmetric policies to showcase how each of these influences the agent's performance. We also show that the learned policies are transferable to different simulation environments. Our results show that SKooP consistently reduces convergence time and increases the learned reward for multiple challenging bipedal locomotion tasks on a quadruped robot. Project page: https://evelyd.github.io/SymmetricKoopmanPredictions