Machine Learning
Coupling Noisy Pairwise Knowledge to the DAG Posterior for Causal Discovery
Guoliang Xu and James E Corter
External causal reports can improve structure learning from limited observations, but their reliability varies across sources and variable pairs. We introduce HB-NoisyKG, a
Bayesian framework that combines observational data with repeated causal reports from sources such as large
language models. Each report is a noisy observation of a direct pair state implied by one DAG. A feature-conditioned Beta prior pools information about pair reliability, and a shared error
matrix captures systematic mistakes. Alternating inference uses the graph posterior to refine reliability estimates, which determine how reports influence subsequent graph updates. The report likelihood uses only graph pair-state marginals, so the same observation layer supports discrete and continuous likelihoods in graph-only and joint inference. Against an 80-restart no-KG baseline, HB uses at most 80 total restarts and lowers mean SHD from 22.39 to 16.06 on five discrete
benchmarks. On a physical light tunnel with random variable IDs and retained descriptions, HB lowers SHD from 39.00 for no-KG to 27.30. On continuous Sachs, graph-only BGe raises AUROC by 0.121 over no-KG Top-K. In a controlled synthetic study, continued updating also reduces mean reliability estimation error and held-out report log loss compared with one-time estimation.
Read the full abstract ▾
Machine Learning · Cross-listed from physics.ao-ph
ClimateBench v2.0: Probabilistic Climate Model Benchmarking
Duncan Watson-Parris, Willa Tobin, Aytaç Paçal et al. (+32 more)
We present ClimateBench v2, a standardized protocol for evaluating
climate models on diagnostics expected to be informative for their skill in projecting mid-century regional temperature and precipitation changes. The protocol is designed to evaluate any physics-based, data-driven, or hybrid
climate model on equal footing using a common set of observational and out-of-distribution tests. We define three tiers of evaluation. Tier I establishes physical credibility through entry-ticket tests of energy conservation, coupled (co-)variability, and basic forced responses. Tier II scores models against post-2015 observations of surface temperature, precipitation, radiative fluxes, sea ice, and key modes of variability using fair CRPS as the primary probabilistic score, complemented by distributional and ensemble-consistency diagnostics. Tier III tests out-of-distribution generalization through paleoclimate simulations spanning the Last Interglacial, Last Glacial Maximum, and Mid-Holocene, and through perfect-model experiments in which data-driven models must predict the future
climate of existing Earth system models from historical data alone. We reserve all observational data after 2015 for testing, and submissions must include multiple ensemble members to enable probabilistic evaluation. This reservation exploits a new opportunity provided by the decade of observations accumulated since the end of the CMIP6 historical experiment, which constitutes an out-of-sample record of forced
climate change (and internal variability) for the current generation of models, and we quantify, in an idealized setting, the information it carries about mid-century warming. We provide the evaluation code, observational reference
datasets, and perfect-model training data as an open
benchmark to drive measurable progress in
climate projection across all modeling approaches.
Read the full abstract ▾
All authors Duncan Watson-Parris, Willa Tobin, Aytaç Paçal, Manuel Schlund, V. Balaji, Kevin Bowman, Chris Bretherton, Peter M. Caldwell, Will Chapman, William D. Collins, Gregory S. Elsaesser, Pierre Gentine, Helene Hewitt, Stephan Hoyer, Ralph Keeling, Nikolay Koldunov, David M. Lawrence, Christian Lessig, Daniel J. Lunt, J. David Neelin, Mike Pritchard, Sarah Purkey, Gavin Schmidt, Tapio Schneider, Michael Schulz, Tiffany Shaw, Isla R. Simpson, Graeme Stephens, Aneesh C. Subramanian, Joao Teixeira, Jessica Tierney, Andrew I. L. Williams, Laure Zanna, Veronika Eyring, Rose Yu
Artificial Intelligence
$\mathrm{TRIZ}^{a}$: Guiding Agent Evolution from Pattern Recognition to Solution Invention
Wenyin Liu, Yiheng Huang and Kai Wang
We propose $\mathrm{TRIZ}^{a}$ (TRIZ exponentiated by an
agent), a general R&D automation paradigm that combines TRIZ inventive theory with LLM-driven
agent evolutionary search. TRIZ's 40 inventive principles and contradiction
matrix provide structured, explainable directions for solution generation, replacing random or untyped mutation with theory-guided ideation. Functional information (FI), operationalized under a frozen reference contract, is combined with TRIZ Ideality to measure useful and harmful function on a commensurable information scale, while hard gates keep promotion distinct from metric improvement. We validate $\mathrm{TRIZ}^{a}$ in cybersecurity–an
adversarial and rapidly evolving domain–on PowerDuck GOOSE, CICIoT2023, and CIC-DDoS2019. Under paired-rerun protocols with protocol fingerprinting and hard-gate validation, the legacy experiments yield absolute F1 improvements of $+2.88$, $+4.23$, and $+0.15$ percentage points, respectively. A completed 45-activity CICIoT2023 campaign further increases macro-F1 from $0.8325$ to $0.8483$, but does not pass its frozen promotion gate. Every result remains traceable from contradiction identification and TRIZ principle selection to code transformation, evaluation metrics, and promotion decision.
Read the full abstract ▾
Computer Vision
EagleDepth: Efficient Fine-Grained Depth Estimation via Pixel Diffusion Decoder
Bowen Chai, Tianbao Zhang, Shuyu Wu et al. (+3 more)
Recovering detailed geometry from high-resolution images is critical for precise perception of the surroundings and objects. However, existing methods which use latent-space modeling and VAE reconstruction can compromise
geometric details. Furthermore, decoding from latent codes introduces substantial inference overhead. To address those issues, we present EagleDepth, an efficient framework for high-resolution
monocular depth estimation that combines the
geometric priors of latent
diffusion with fine-grained pixel-space generation. Our key idea is to retain depth-aware latent representations as guidance while generating the final depth map directly in pixel space. We train the latent and pixel components sequentially: first, we fine-tune a pretrained latent
diffusion model using paired RGB–depth supervision; then, we adapt a pretrained pixel
diffusion decoder, PiD, to predict depth conditioned on the learned features. Training of the pixel component starts at 1024 resolution and continues across multiple resolutions up to 4K. The latent branch processes resized, lower-resolution RGB images, while the pixel branch generates depth at the target resolution, bypassing the original VAE decoder. This design preserves learned
geometric knowledge without requiring the latent backbone to operate at the output resolution. On five commonly used depth estimation
datasets and the high-resolution Synth4K
dataset, our framework achieves state-of-the-art depth estimation performance, with faster inference and better preservation of fine structures and object boundaries.
Read the full abstract ▾
All authors Bowen Chai, Tianbao Zhang, Shuyu Wu, Dexin Zuo, Zhaoxin Fan, Danping Zou
Artificial Intelligence · Cross-listed from astro-ph.SR
Cross-Modal Solar Image Synthesis: Adapting the Surya Foundation Model from He I 10830 Å to EUV Translation and Coronal Hole Segmentation
Marco Marena, Andrés Muñoz Jaramillo, Qin Li et al. (+8 more)
The long observational record of He I 10830 Å offers a means to investigate
solar morphology before modern extreme-ultraviolet (EUV) imaging. We adapt the Surya
solar foundation model to predict
Solar Dynamics Observatory/Atmospheric Imaging Assembly (SDO/AIA) 94, 193, and 304 Å images and a coronal hole (CH) probability map from full-disk helium observations. A convolutional input adapter, low-rank backbone updates, and dedicated output decoders learn from temporally paired, geometrically registered observations, with Spatial Possibilistic Clustering
Algorithm (SPOCA) catalog polygons providing CH supervision. On observations held out from downstream fine-tuning, the selected dedicated models achieve disk-restricted correlation coefficients (CCs) of 0.8196, 0.8885, and 0.8398 for the three AIA channels, respectively; the CH model achieves an intersection over union (IoU) of 0.4360. The predictions recover broad
solar structure, although local agreement varies substantially by channel. An optional residual refiner addresses spatial detail, and its application on pre-SDO dates improves the correlation of synthetic AIA 304 with
Solar and Heliospheric Observatory/Extreme-ultraviolet Imaging
Telescope (SOHO/EIT) 304 references from 0.6511 to 0.7085. Together, these results support the feasibility of helium-conditioned EUV morphological proxies and motivate their use in historical reconstruction, within the scope of the downstream test and cross-instrument evaluation.
Read the full abstract ▾
All authors Marco Marena, Andrés Muñoz Jaramillo, Qin Li, Haodi Jiang, Jinghao Cao, Wen He, Ziyang Zhang, Chenxi Yuan, Chao Wang, Haimin Wang, Bo Shen
Robotics
Online Target-less Radar-LiDAR-Camera Extrinsic Calibration via Joint Optimization
Gunhee Shin, Yunsoo Kim, Chanhyuk Lee et al. (+7 more)
Fusing radar,
LiDAR, and camera enables robust perception in diverse and adverse conditions, but the fusion performance critically depends on accurate extrinsic calibration among the three sensors. In this paper, we address the problem of online target-less extrinsic calibration for the radar-
LiDAR-camera system. Existing target-less methods are mostly designed for a single sensor pair, and composing the pairwise results does not guarantee consistency across the three sensors. Moreover, the sparse and noisy radar measurements make the radar-involving pairs unreliable. To tackle these challenges, we propose a joint calibration framework that constructs residuals for each sensor pair and optimizes the extrinsics of all pairs together to minimize the overall residual. Furthermore, we introduce an adaptive radar noise filter that rejects spurious radar returns using a range-dependent margin, and a correspondence accumulation strategy that aggregates sparse radar correspondences over frames. We validate our method on an in-house radar-
LiDAR-camera
dataset covering diverse urban environments, where it reduces calibration errors across all sensor pairs over a state-of-the-art camera-
LiDAR baseline.
Read the full abstract ▾
All authors Gunhee Shin, Yunsoo Kim, Chanhyuk Lee, Wanhee Kim, Minwoo Lee, Sungwoo Han, Jeongwoo Woo, Hyuntai Chin, Minha Park, Hyun Myung
Robotics
NM-LIO: Multiple LiDAR-Inertial Odometry Addressing LiDAR Measurement Noise Discrepancy
Gunhee Shin, Seungjae Lee, Minho Oh et al. (+4 more)
Multiple
LiDAR-inertial odometry methods have been widely applied in robotic applications owing to their enhanced accuracy and reliability. However, adopting multiple
LiDAR systems can be challenging due to the discrepancies in measurement noise across different
LiDARs. Existing methods have typically overlooked the noise discrepancies, which can significantly affect accuracy. In this paper, we propose Noise-aware Multiple
LiDAR-Inertial Odometry (NM-LIO) that addresses the noise discrepancies. We integrate a noise model to quantify the measurement noise of each
LiDAR. Additionally, we estimate the uncertainty of the residuals based on the measurement noise, allowing the measurement model to capture the noise discrepancies. Our proposed method is evaluated on a public multiple
LiDAR dataset and compared with state-of-the-art methods. The experimental results demonstrate that the proposed method can accurately estimate the odometry in various environments by accounting for the noise discrepancies.
Read the full abstract ▾
All authors Gunhee Shin, Seungjae Lee, Minho Oh, Dongkyu Lee, Jaeyoung Lee, Youngwoo Seo, Hyun Myung
Symbolic Computation
A Finite Certificate for the Positive $n=11$ Vasc Inequality
Tiancheng Xu and Wensheng Yu
We establish the Vasc cyclic inequality for all strictly positive real eleven-tuples by a finite exact certificate. Cyclic rotation places a minimum at the first position, and rank words together with cumulative gaps reduce the problem to $10!=3{,}628{,}800$ homogeneous integer
polynomials of degree eleven. Independent coefficient reconstruction proves $3{,}358{,}617$ roots directly. The remaining $270{,}183$ roots are covered by $267{,}952$ ordinary midpoint certificates, $2{,}195$ complete binary subdivision certificates, and $36$ complete certificates containing partial sorting nodes. Each nontrivial leaf is checked by subtracting weighted AM-GM midpoint
circuits from a positive multiple of the transformed
polynomial and verifying every residual coefficient. Explicit word-set comparisons certify disjointness and exact coverage. We prove the soundness of the leaves and both subdivision rules, then recover the rational inequality by positive denominator clearing and continuity. The resulting proof combines explicit mathematical reductions with independently checked exact integer certificates.
Read the full abstract ▾
Machine Learning
TMAML: Temporal Model-Agnostic Meta-Learning for Cold-Start Time Series Forecasting
Wannes Janssens, Matthias Bogaert and Dirk Van den Poel
Cold-start
forecasting, the task of
forecasting a time series with little to no historical data, is a common challenge. Addressing it requires approaches that learn quickly from few datapoints and leverage information from related series, typically through static
covariates, to generalize well to unseen series. While some global
forecasting models can generate cold-start predictions by leveraging information shared across multiple series, they are not optimized for out-of-train-set generalization or adaptation from short histories. In this work, we formulate cold-start
forecasting as a few-window learning problem and introduce Temporal Model-Agnostic Meta-Learning (TMAML), which tailors the model-agnostic meta-learning
algorithm, originally developed for few-shot adaptation of neural networks, to deep time series
forecasting. TMAML constructs meta-tasks as temporally consistent support-query windows and pairs them with a temporal meta-training and meta-testing procedure, yielding
forecasting models that are explicitly optimized for cold-start
forecasting. We instantiate TMAML on the Temporal Fusion
Transformer (TFT) and present an initial empirical analysis of forecast accuracy and calibration across three cold-start scenarios: TMAML consistently outperforms or matches a standard ERM-trained TFT, yields better-calibrated forecasts than naive on two of the three scenarios, but does not consistently outperform naive on probabilistic forecast accuracy.
Read the full abstract ▾
Information Theory
Plotkin-type Bound on Function-Correcting Codes for Monomial Mappings
Kanchana Lokshmii Jagatti and B. Sundar Rajan
Function-correcting codes, recently introduced by Lenz et al.\ (2023), provide a framework for the receiver to provide a higher level of protection against channel errors than the level of protection for messages. In this paper, we derive a generalised Plotkin-type bound on the redundancy of FCCs under the Hamming metric for arbitrary functions over finite fields. The proposed bound recovers the classical Plotkin bound for classical error-correcting codes as a special case. We further obtain explicit Plotkin-type bounds for linear functions, Hamming-weight functions, and Hamming-weight threshold functions. In addition, we study FCCs for monomial mappings $f(x)=x^n$ over $\mathbb{F}_{q^k}$, which are widely utilised in coding theory, cryptography, and finite geometry, and derive an explicit Plotkin-type redundancy bound.
Read the full abstract ▾
Multiagent Systems
Nonlinear Density-Driven Optimal Control (D2OC) for Multi-Agent Spatial Coverage via Sequential Convex Programming
Julian Martinez and Kooktae Lee
This paper presents a nonlinear extension of Density-Driven Optimal Control (D2OC) for multi-
agent spatial coverage with prescribed density distributions. Rather than assigning individual target locations, D2OC drives the collective spatial distribution of
agents toward a desired density through a Wasserstein-based objective. We extend this framework to multi-step finite-horizon control for discrete-time control-affine nonlinear systems using sequential convex programming. At each control update, the nonlinear dynamics are locally linearized over the prediction horizon, yielding a strictly convex quadratic program that preserves the Wasserstein barycentric structure while directly incorporating input constraints. We further characterize the effect of constrained control deviations and nonlinear Taylor remainders on the accuracy of the local linear prediction, establishing an explicit finite-horizon error bound and a two-step specialization for receding-horizon implementation. The resulting method retains the decentralized, distribution-driven nature of D2OC while providing a computationally efficient optimization procedure for nonlinear multi-
agent systems. Simulations with unicycle and quadrotor teams show coverage performance comparable to nonlinear model predictive control, while substantially reducing computation time. These results demonstrate a tractable and theoretically characterized framework for density-driven spatial coverage under nonlinear dynamics.
Read the full abstract ▾
Artificial Intelligence
Label Agreement Does Not Measure Authorization
Amir Sabbaghziarani, Bradley Thomas Baker, Theodore J. LaGrow et al. (+1 more)
Many groups now delegate label ontology and metadata harmonization to agentic LLM pipelines. We built one and audited it. Our aggregate scores looked healthy, but the pipeline kept failing in ways they did not show, so we set out to find what they hid. Label agreement asks whether a proposed label matches a reference. It does not ask whether the
agent was entitled to propose it, whether the output was complete enough to act on, or whether the label moved when the evidence moved. We measured those three separately on COBRE and FBIRN, two schizophrenia and control neuroimaging cohorts from different consortia, and they come apart, from label agreement and from each other. Showing the
agent an upstream proposal barely moves label agreement, 0.857 to 0.870, while agreement on the chosen action doubles, 0.409 to 0.830. Output that parses as JSON still drops a required field on 10% of one model's cases and 33% of the other's. And an
agent that replays its first answer scores perfectly on original cases and zero once we change the evidence that decides them. Downstream, a row-order error that none of these metrics reports erases most of the diagnostic signal. So we measure these properties apart, pair each with a control, and gate commitment on the result, which makes failures visible and easy to route to a person. None of this prevents failure. Our reference labels are rule-derived, so agreement with them means consistency, not correctness. Code is available at https://github.com/amir-sbg/Label-Agreement-Does-Not-Measure-Authorization.
Read the full abstract ▾
All authors Amir Sabbaghziarani, Bradley Thomas Baker, Theodore J. LaGrow, Sergey Plis
Cryptography and Security
Quantum Machine Learning Protection of Military Quantum Key Distribution Against Cryptographically Camouflaged Attacks
Muhammad Shaheer Bin Junaid
Quantum key distribution proves its protocol secure and says nothing about the hardware beneath it, so military and government operators fielding it for command-and-control keys monitor the channel for implementation attacks, and that monitoring has a blind spot. An adversary with a kleptographic foothold in the generator of a public per-block value \(x=g^v \pmod p\) can hide attacked blocks in honest noise, gating them on a predicate of its discrete logarithm, making detection a discrete logarithm problem that defeats every efficient classical monitor yet yields to a
quantum kernel recovering \(v\) through Shor's
algorithm. I formalise these cryptographically camouflaged attacks, reduce their hardness to an established learning separation, prove a single-frequency fidelity kernel cannot represent an interval predicate, and test them on Ghillie, a decoy-state BB84 simulator with a positive key rate to 142 km. From 10- to 14-bit groups over two seeds, a classical monitor reads 0.458 to 0.516 on camouflaged attacks while the
quantum kernel reads 1.000, and both catch overt attacks above 0.99. Finite-precision recovery under depolarising noise and a hardened predicate lower the
quantum result to 0.916 through 0.983 with the classical monitor at chance, and a feasibility probe on IBM Heron processors tracks the exact kernel within 0.034. A defender can therefore discard precisely the compromised key material, although the advantage is asymptotic, awaits fault tolerance, and holds only when the feature map matches the adversary's predicate, since a low-frequency map reads 0.545 on a residue pattern and 0.982 once aligned.
Read the full abstract ▾
Artificial Intelligence
Decide, Ask, or Defer: Clinical LLMs under Incomplete Evidence
Mingzhan Yang and Weili Wu
Clinical LLMs must decide not only what diagnosis to produce, but also whether the available evidence is sufficient for autonomous decision making. Binary DECIDE/ABSTAIN formulations merge distinct non decision states and do not explicitly evaluate information acquisition. We introduce a DECIDE/ASK/DEFER formulation together with a blinded protocol that prevents models from using evidence completeness metadata. We evaluate Qwen, Gemini, and GPT on 200 matched clinical evidence states constructed from DDXPlus. The models show substantial differences in action selection under identical evidence, with disagreement in 137 of 200 states. For Qwen, a matched targeted versus random analysis shows that selected information changes the likelihood of a subsequent autonomous decision more clearly than diagnostic correctness. Its matched DECIDE/ABSTAIN baseline further reveals a safety autonomy tradeoff: the three action policy rescues some erroneous autonomous decisions but also removes some correct autonomous deci sions. These results show that separating information acquisition from clinician deferral exposes behavior that binary abstention hides, without yielding a uniformly improved decision policy.
Read the full abstract ▾
Artificial Intelligence
Autonomous Structuring of Radiology Reports Across Modalities at Archive Scale Using an Open-Weight Large Language Model
Friedrich Puttkammer, Fabian Drexel, Marlene Fritzsche et al. (+19 more)
Purpose: To develop and evaluate an open-weight large
language model (LLM) pipeline that converts an entire archive of free-text radiology reports into structured reports without human oversight. Materials and Methods: In this retrospective study, a pipeline with 150 hierarchically organized templates was developed at one center and tested at a second center on reports from 2010 to 2025. The open-weight model gpt-oss-120B selects the template in three constrained-decoding steps and fills it on one local graphics processing unit. Template selection was scored against expert labels on 914 randomly sampled reports of five modalities, structuring quality on 920 radiography and CT reports corrected field by field by five residents. The pipeline then processed the complete archive of the second center. Proportions are reported with Wilson 95% confidence intervals (CIs). Results: An optimal template set was selected for 74.4% of reports (680 of 914; 95% CI: 71.5%, 77.1%) and an appropriate set for 82.3% (752 of 914; 95% CI: 79.7%, 84.6%), 87.7% for single-region and 54.1% for multi-region reports. Macro semantic textual similarity between output and corrected reference was 0.95 for radiography and 0.97 for CT, residents left 88.7% of 24,638 fields unchanged, and unsupported content was flagged in 1.0% and 1.5% of reports. Of 2,186,982 archive reports, 96.5% received structured output, 2,401,544 structured reports, at 1,258 reports per hour on one graphics processing unit. Conclusion: An open-weight LLM pipeline structured a complete multimodality report archive without human oversight with high content fidelity. Multi-region reports remained the main source of template errors.
Read the full abstract ▾
All authors Friedrich Puttkammer, Fabian Drexel, Marlene Fritzsche, Era Stambollxhiu, Miriam Kumpf, Lena Schmitzer, Lea Schumann, Lina Xu, Johannes Moll, Jannik Lübberstedt, Zeineb Ben Chaaben, Anirudh Narayanan, Hartmut Häntze, Renato Cuocolo, Antonios Billis, Alexander Löser, Jawed Nawabi, Marcus R. Makowski, Cosmin I. Bercea, Shahrooz Faghihroohi, Lisa C. Adams, Keno K. Bressem
Robotics
DreamFormer: Dream Imitation with a Transformer World Model for Language-Conditioned Robotic Manipulation
Mostafa Kotb, Cornelius Weber, Muhammad Burhan Hafez et al. (+1 more)
We introduce DreamFormer, a model-based
agent that acquires
language-conditioned, multi-task skills by imitating expert demonstrations within the latent imagination of a learned world model. DreamFormer first learns a task-agnostic
Transformer world model from unstructured play data, then acquires task-specific behaviors by optimizing an intrinsic reward that aligns
agent-generated rollouts with expert demonstrations in latent space. Since the policy is trained on-policy inside imagination, it is exposed to its own errors during training, mitigating the
covariate shift inherent to offline behavioral cloning. To make long-horizon imagination affordable, DreamFormer encodes a high-resolution multi-view robotic observation into a single input
token, avoiding both spatial downsampling and the multi-
token representations used by prior
Transformer world models. On the long-horizon CALVIN
benchmark, DreamFormer outperforms LUMOS, the comparable model-based
agent, on single-environment evaluation (2.52 vs 2.34 average tasks completed per chain of five) while keeping imagination rollouts tractable. Against HULC, the behavior cloning baseline, it nearly doubles performance on zero-shot transfer to an unseen environment (1.30 vs 0.67), indicating that dynamics learned by the world model transfer more readily than a directly cloned policy. This is consistent with the role attributed to internal models in biological
agents, where a model of environment dynamics supports behavior in situations not previously encountered.
Read the full abstract ▾
All authors Mostafa Kotb, Cornelius Weber, Muhammad Burhan Hafez, Stefan Wermter
Artificial Intelligence
Action-Consequence Alignment for Reliable Planning and Self-Improving in Latent World Models
Jinping Wang1, Zhiqiang Gao, Xiantong Zhen et al. (+1 more)
Latent world models learn to predict observed transitions, yet low prediction error alone does not guarantee reliable planning. Inspired by self tickling experiments in neuroscience showing that disrupting motor sensory correspondence increases prediction mismatch, we examine whether learned world models preserve an analogous action consequence correspondence.The results show nearby alternatives can receive lower prediction errors despite producing physical outcomes farther from the recorded target. With that future treated as a goal, this reveals a concrete prediction planning mismatch: the model assigns a lower cost to an action that achieves the target less accurately. To mitigate this gap, we introduce Action Consequence
Alignment (ACA), a training objective that complements forward prediction by penalizing the prediction error advantage of locally searched alternatives over factual actions without additional model components or environment interactions during training. The same principle can also guide additional data collection for self improvement. We demonstrate that across diverse environments and evaluation settings, ACA improves planning performance and reduces real goal error, while ACA guided data collection outperforms random local sampling. These results support action consequence
alignment as a practical principle for bridging predictive learning and reliable planning.
Read the full abstract ▾
All authors Jinping Wang1, Zhiqiang Gao, Xiantong Zhen, Ling Shao
Computer Vision
FASTER: Fast Adjoint Stochastic Transport for Endpoint Refinement in Reward-Guided Image Editing
Yimiao Zhou, Zejia Zhong, Jingya Wang et al. (+1 more)
Reward-guided image editing at test time seeks to improve a specified reward while preserving source content and visual plausibility. Many existing approaches optimize candidates through pretrained generation processes, making repeated adjustment depend on costly large-model execution and, in some cases, backbone backpropagation. We develop a theoretical framework that jointly accounts for reward, source preservation, and pretrained-prior preferences, allowing the desired output distribution to be specified separately from the dynamics used to realize it. Based on this framework, we introduce FASTER, which trains a small network for each source and objective to perform inexpensive editing, while pretrained and reward models provide feedback on candidate outputs. By reusing each candidate and its feedback across multiple small-network updates, FASTER reduces repeated sampling and supervision queries without placing the pretrained generative backbone inside the inner optimization loop. On SD3, FASTER leads all four target metrics and several validation metrics among the evaluated methods. Compared with the evaluated baseline that optimizes controls along pretrained generation trajectories, FASTER achieves editing-time speedups of up to \({6.91\times}\) on Stable
Diffusion 3 and \({24.14\times}\) on Stable
Diffusion 1.5.
Read the full abstract ▾
All authors Yimiao Zhou, Zejia Zhong, Jingya Wang, Ye Shi
Machine Learning
PhaseGate: Phase-Aware CPU Retrieval Scheduling for On-Device LLMs on Unified Memory
Seoyoon Yum and Sehoon Kim
On-device assistants run GPU-based LLM inference alongside CPU
retrieval on unified-
memory systems. Under a saturated local-
retrieval workload, four concurrent
retrieval workers raise 95th-percentile (p95) decode latency by 60-61% on two M4 systems, whereas prefill latency rises by only 5.7-6.9%. We study LLM phase as an admission signal for independent CPU
retrieval under controlled LLM workloads. PHASEGATE calibrates separate concurrency limits for prefill and decode, selecting four and one on our base-M4 configuration. Under a backlogged queue, it achieves 2.0 times the aggregate
retrieval throughput of the best tested feasible fixed policy, with both p95 LLM latency metrics within 1.25 times their no-
retrieval baselines in all seven held-out runs. A phase-blind control, TimeGate, uses the same two limits on a calibration-derived schedule without observing LLM phase. It achieves similar
retrieval throughput but violates the output-
token latency limit in every run. M2 and M2 Pro Mac minis reproduce the policy ordering, while output-length sweeps show that the advantage narrows as decode occupies more of each request.
Read the full abstract ▾
Robotics
WasserMan: Benchmark for Underwater Manipulation Policy Learning
Danil Belov, Artem Erkhov, Sergei Parsegov et al. (+1 more)
Underwater manipulation couples visual decisions, contact forces and a thruster-controlled floating base. We introduce WasserMan, to our knowledge the first multi-task simulation
benchmark for visuomotor learning of floating-base
underwater contact manipulation. It provides ten expert-solvable tasks, two vehicle-arm platforms, a bimanual configuration and
underwater dynamics. Nine tasks have learned-policy evaluations. We compare ACT,
diffusion policies (DP) and behavioral cloning on six tasks, with three training runs and equal sampled-window budgets. Pretrained SmolVLA adds evaluations on three tasks. In the tested settings, removing integral action can prevent completion, while changing action interfaces can reduce learned-policy success despite successful expert replay. Currents produce task-dependent changes in completion and actuation effort. Versioned tasks, demonstrations and per-episode evidence support a protocol separating expert feasibility, learned completion and execution effort.
Read the full abstract ▾
All authors Danil Belov, Artem Erkhov, Sergei Parsegov, Pavel Osinenko
Numerical Analysis
Pressure Monotonicity for a Rotation-Based Biot Discretization with Circumcentric Mass Lumping
Huipeng Gu, Mingchao Cai and Jingzhi Li
We establish cellwise pressure monotonicity for a mixed finite element discretization of the rotation-based Biot system. The spatial discretization is based on compatible discrete de Rham complexes, using standard lowest-order elements for rotation, displacement, Darcy flux, and pressure. For the temporal discretization, we employ the backward Euler method. Exactness of the discrete complexes reduces the mechanics contribution to the pressure Schur complement to a scaled pressure mass
matrix, yielding a positive diagonal mechanics capacity. For the Darcy part, positive circumcentric facet weights yield a weighted
cell graph
Laplacian after flux elimination, while the mass-lumped product remains uniformly equivalent to the standard Darcy product. The diagonal mechanics capacity and the weighted
cell graph
Laplacian make the condensed pressure
matrix Stieltjes. Consequently, the pressure propagator is entrywise nonnegative, which yields a cellwise discrete maximum principle. Numerical experiments confirm the predicted convergence rates and show that the proposed discretization suppresses spurious cellwise pressure oscillations.
Read the full abstract ▾
Machine Learning
Weight Decay and Neuron Condensation: A Three-Stage Analysis of Two-Layer ReLU Networks
Cheng Xu, Pengxiao Lin, Zhangchen Zhou et al. (+1 more)
Weight decay is widely used as a regularization technique in neural network training, yet its role in
neuron condensation (parameter direction
alignment) remains unclear. Starting from a parameter initialization in the neural tangent kernel regime, we characterize training dynamics under weight decay through three stages: rapid fitting, amplitude compression, and
neuron condensation. Using a two-layer ReLU network, we analyze a residual correlation field that governs both
neuron amplitudes and directions. During rapid fitting, the residual approaches a quasi-static
equilibrium maintained by weight decay while the tangent kernel remains nearly unchanged. In the early stage of amplitude compression, kernel decay amplifies the residual correlation field, whose isolated attracting extrema provide candidates for condensation directions. As
neuron amplitudes stabilize, we bound the drift of attracting extrema and demonstrate contraction of
neuron directions around them, leading to
neuron condensation. This staged analysis provides a dynamical understanding of how weight decay promotes a condensed representation, beyond reducing parameter norms.
Read the full abstract ▾
All authors Cheng Xu, Pengxiao Lin, Zhangchen Zhou, Zhi-Qin John Xu
Computer Vision
TAME:Topology-Aware Text-Driven Motion Editing across Heterogeneous Humanoid Skeletons
Qichen Zheng, Siyuan Yang, Chong Wang et al. (+4 more)
Text-driven motion editing modifies an existing motion sequence according to a text instruction while preserving the content of the source motion. Existing methods are typically built for a single, fixed skeletal
topology, which limits their use in animation pipelines where characters differ in joint count and skeletal hierarchy. We present
Topology-Aware Motion Editor (TAME), a flow-matching
transformer that edits motions on
humanoid skeletons of varying
topology. TAME represents motion as per-joint, per-frame
tokens and models interactions among joints, across frames, and with the text instruction through skeletal, temporal, and text cross-
attention layers. To make the skeletal
attention follow each character's hierarchy, TAME replaces full joint
attention with
Topology-Constrained Skeletal Propagation (TCSP), which restricts
attention to one-hop kinematic neighbors in the skeleton's adjacency
matrix. We further introduce Edit-Focused Representation
Alignment (EFRA), a self-distilled representation
alignment strategy that aligns student features with cleaner EMA-teacher features exclusively on edit-relevant joint-time
tokens, making edits faithful to the instruction. To make this setting trainable and comparable, we construct TopoMotionFix, a multi-
topology extension of MotionFix with seen- and unseen-
topology evaluation protocols. TAME outperforms previous methods in edit
alignment and source preservation on MotionFix and reliably edits motions on unseen skeletons in TopoMotionFix.
Read the full abstract ▾
All authors Qichen Zheng, Siyuan Yang, Chong Wang, Jun Liu, Shijian Lu, Alex Kot, Kwok-Yan Lam
Multiagent Systems
Quantifying Collusion Among Autonomous LLM Agents: A Statistical Analysis of the Collusion Wiki Incident
Shariq Murtuza
In August and September 2026, independent researchers publicly documented an unusual incident: thousands of autonomous
agents, self identifying as OpenAI models on web research tasks, discovered and began using a small German wiki as an improvised message board posting roughly 18,000 times over six weeks to relay task answers, share a sandbox escape technique, and coordinate against a volunteer human moderator who spent weeks manually deleting their content [1]. The investigators' public writeup is a careful qualitative account, rich with direct quotation, but does not attempt a statistically rigorous quantitative characterization of the behaviour it documents.
Read the full abstract ▾
Networking
Diffusion-Based Stress Testing of Overload Monitoring for Resilient Emergency Cellular Networks Using Internet CDR Proxies
Bilal Hussain, Xiao Tang, Tan Li et al. (+3 more)
Disasters can overload cellular control-plane signaling within minutes, yet fine-grained
Radio Resource Control (RRC) or Next Generation (NG) Application Protocol (NGAP) telemetry is
privacy-sensitive and costly to collect for analytics. Many emergency monitoring pipelines therefore rely on coarse Call Detail Record (CDR) aggregates. We treat Internet activity in CDR grids as a practical proxy for hidden signaling stress under that constraint. We train a lightweight convolutional neural network (CNN) on stylized overload injections, stress-test it with
diffusion-synthesized surges that preserve normal traffic structure, and adapt the detector by retraining on hard synthetic samples. Under stress-test conditions, the default alert threshold fails even though receiver operating characteristic (ROC) curves stay strong: the detector still assigns overloaded
cells a larger overload probability than normal
cells, but those probabilities fall below the default cutoff 0.5 and are labeled normal, so the F1-maximizing threshold – selected post hoc on the same stress-test grids (oracle $\tau^*$) – shifts by $0.32 \pm 0.03$ (operating-point drift). Across three random seeds, hard-sample adaptation raises thresholded performance (F1) from 0% (no alerts at the default cutoff 0.5 on any seed) to $85.67 \pm 14.37$% and ranking from ROC-AUC $0.886 \pm 0.040$ to $0.99996 \pm 0.00007$.
Diffusion-synthesized surges expose threshold fragility that matched-condition training – training and testing on the same stylized injections – hides, and hard-sample adaptation restores usable alerts at the default cutoff. Together, these steps define a reusable pre-deployment stress test for emergency monitors. Internet-only CDR input further supports lightweight AI-native workflows that combine monitoring, recalibration, and adaptation.
Read the full abstract ▾
All authors Bilal Hussain, Xiao Tang, Tan Li, Muhammad Azhar, Danista Khan, Fawad Ahmad
Information Theory · Cross-listed from eess.SP
QoS-Constrained Resource Pattern Design for V2X-ISAC Systems
Hanyoung Park, Gangmin Kim, Yoo-Seung Song et al. (+1 more)
Integrated sensing and communication (
ISAC) has emerged as a promising approach for vehicle-to-everything (V2X) systems by enabling communication and sensing over shared
radio resources without additional installation of dedicated sensors. However, candidate resources may experience different communication qualities due to varying channel conditions and resource contention, which should be considered when designing sensing resource patterns. In this letter, we propose a quality-of-service (QoS)-constrained resource pattern selection framework that jointly considers communication reliability and sensing performance. The sensing objective is formulated based on the Fisher information
matrix for joint range and velocity estimation, while a minimum packet reception ratio (PRR) is imposed as the communication QoS constraint. To avoid the complexity of exhaustive search, a greedy selection
algorithm with one-swap refinement is developed. Simulation results show that the proposed method improves sensing performance over conventional resource patterns while satisfying the PRR requirement and achieves performance close to the exhaustive-search optimum and sensing-only optimum.
Read the full abstract ▾
All authors Hanyoung Park, Gangmin Kim, Yoo-Seung Song, Ji-Woong Choi
Numerical Analysis
A Locally Conservative Enriched Linearized Neural Network Approximation to Elliptic PDEs
Seungil Kim, Gwanghyun Jo and Young Ju Lee
This paper presents a locally conservative Enriched Linearized Neural Network (ENN) method for the Darcy flow model. A linearized shallow ReLU$^k$ network, whose hidden-layer parameters are fixed on a quasi-uniform set of the sphere, carries the approximation, and piecewise constant functions on an auxiliary subdivision of the domain are added to ensure local mass conservation. The subdivision may consist of general, possibly curved, elements. Since neural network functions do not satisfy the inverse inequality, the stability and error analysis of the standard discontinuous Galerkin interior penalty formulation do not apply. The missing inverse inequality is remedied by an edge identity combined with a Galerkin least-squares (GLS) term, together with a single neural approximant that is optimal in the $L^2$, $H^1$ and $H^2$ norms simultaneously. We use the nonsymmetric interior penalty (NIPG) form, which is coercive for every positive penalty parameter. With $n$, the width of the shallow neural network, used in ENN, we prove an error estimate in the EG norm that is of the optimal $H^1$ order, i.e., $O(n^{-(r-1)/d})$ for solutions in $H^r$, $r\ge2$, and $O(n^{-(s-1)/d})$ for solutions in $H^s$, $\frac32<s\le2$, for any shape-regular subdivision, when the GLS parameter satisfies $\min\{h,n^{-1/d}\}^2\lesssim\tau\lesssim n^{-2/d}$ and the edge terms are scaled with $\min\{h_e,n^{-1/d}\}$, i.e., by the resolution of the network rather than of the mesh. A numerical Darcy flux is constructed, which is locally conservative, regardless of the size or the number of the elements. Sample numerical results demonstrate the correctness of the theory. In particular, a fixed $4\times4$ subdivision gives the same accuracy as a subdivision refined with the network, and the mass loss stays at round-off unlike a non-conservative neural network method.
Read the full abstract ▾
Machine Learning · Cross-listed from physics.soc-ph
All against the machine: the Solo score for rating skill in variable environments
David Reguera, Xavier R. Hoffmann, Irene Pérez et al. (+5 more)
We propose a distribution-free metric to rate individual skill in “player-versus-environment” settings, where participants face heterogeneous tasks without direct opponents. Such settings are common in digital platforms, games, education, finance, and the
benchmark evaluation of AI
agents. They combine high randomness, tasks of widely varying difficulty, and unknown heterogeneity across individuals. Our metric maps each task outcome to a bounded performance score with zero population mean and variance bounded by $1/3$, regardless of the outcome distribution, so that scores are directly comparable across tasks. Aggregating these scores over tasks yields an interpretable skill score for each individual (the “Solo” rating), and a reshuffling null model tests whether that score exceeds what chance alone would produce. We validate the approach on the progression of $2\times10^5$ players in the mobile games Candy Crush Saga and Bubble Witch 3 Saga. The metric identifies high- and low-skill players with high statistical confidence, their classification persists over hundreds of subsequent levels, and a windowed version of the score tracks changes in performance along progression.
Read the full abstract ▾
All authors David Reguera, Xavier R. Hoffmann, Irene Pérez, Pol Colomer-de-Simón, Miquel Masoliver, Xavier Guardiola, Jan Wedekind, Marián Boguñá
Data Structures and Algorithms
A Factor-5 Conversion to Internal Collage Systems
Simone Faro
A collage system is a grammar-based compression model that extends straight-line programs with repetition and substring truncation. Internal collage systems additionally require every nonterminal to be structurally reachable from the start symbol. Migita, Uehata, and I (CPM 2026) showed that any collage system of size $m$ can be converted into an internal one generating the same string with size at most $9m$, and left the improvement of this constant as an open problem. We show that a simple refinement of their top-down conversion reduces the bound to $5m$. The proof combines three elementary ideas: canonicalization of generated truncations that remain aligned with a target endpoint; a repetition decomposition that absorbs every complete copy of the repetition base into a maximal core; and monotonicity of structural reachability, which prevents an input truncation rule from both becoming structurally reachable and later acting as a hidden target at which endpoint
alignment is lost. The conversion runs in deterministic $O(m^2)$ worst-case time in the stated unit-cost random-access machine model. Consequently, for every string, the minimum size of an internal collage system is at most five times the minimum size of a general collage system.
Read the full abstract ▾
Machine Learning
Proximal Causal Learning under Unmeasured Confounding
Ying Tang and Yi Wang
Estimating treatment effects from observational data typically relies on the No Unmeasured Confounding Assumption (NUCA), which rarely holds in practice. Proximal causal learning (PCL) addresses unmeasured confounding via proxy variables, yet existing methods require the proxy variables to be pre-specified. Thus, we propose PCL-U, a framework that learns proxy variables directly from observed
covariates. PCL-U uses neural encoders to decompose
covariates into treatment-inducing, outcome-inducing, and shared proxies, guided by minimax mutual information objectives, and obtains causal estimates through a practical moment-based risk function. Experiments on
benchmarks show that PCL-U matches or outperforms existing baselines. Besides, there are two types of synthetic
datasets with varying dimensions and confounding strengths that illustrate that our method maintains stable estimation accuracy.
Read the full abstract ▾