Machine Learning
Reinforcement Learning with a Bilevel World-Model Architecture for Scan-Order Optimisation in Laser Directed Energy Deposition
Xian Wu, Haoran Li, Yuanqi Chu et al. (+2 more)
Scan-order design in
laser directed energy deposition (LDED) is a delayed, path-dependent thermo-mechanical decision problem, because sequence quality becomes observable only after the complete deposition and cooling cycle. This work formulates LDED scan-order optimisation as a finite-horizon, permutation-constrained reinforcement-learning problem and develops a bilevel finite-element-teacher-labelled AI workflow. A surrogate-assisted teacher-guided optimisation loop learns the Abaqus-labelled response landscape and provides a tractable terminal-reward environment for policy training. A frozen Maskable Proximal Policy Optimization (MaskablePPO) policy is then used to generate legal scan-order candidates, which are independently validated through Abaqus thermo-mechanical simulations. The results show bounded, N-dependent policy-generation value rather than record-level dominance over the mature surrogate-assisted optimiser. The strongest scan orders are obtained by the teacher-guided surrogate loop, whereas PPO autonomously reaches competitive regions of the native response landscape, with stronger rank concentration at smaller track counts and a clear reliability boundary at longer horizons. The teacher-labelled landscape further supports a physically gated lexicographic reward hierarchy in which warpage admissibility is the primary constraint, plastic strain acts as a safety filter and residual-stress-related improvement is pursued conditionally within the admissible region. Validated sequences also reveal an interpretable scale-separated ordering tendency that combines global spatial dispersion with local structured grouping. This workflow provides a route from fixed scan-rule selection toward finite-element-teacher-validated policy generation, while preserving independent finite-element validation as the final physical gate.
Read the full abstract ▾
All authors Xian Wu, Haoran Li, Yuanqi Chu, Dongbin Zhao, Bin Wang
Machine Learning
GL-LFGNN:A Global-Local Dual-branch Causal Graph Neural Network Based on Liang-Kleeman Information Flow for EEG Emotion Recognition
Ziyi Wang and Dongyang Kuang
EEG-based
emotion recognition holds significant promise for objective diagnosis of mood disorders. Graph neural networks (GNNs) have emerged as the dominant paradigm for modeling inter-channel dependencies in
EEG, yet existing approaches rely on symmetric adjacency
matrices derived from spatial proximity or functional correlations that fundamentally capture statistical associations rather than directed causal influences, which conflicts with the inherently asymmetric, causally-driven nature of neural information flow. To bridge this gap, we propose GL-LFGNN, a Global-Local Dual-branch Causal Graph Neural Network grounded in Liang-Kleeman information flow theory. Unlike Granger causality that merely assesses temporal precedence, our approach rigorously quantifies causal strength from a dynamical systems perspective, yielding neurophysiologically interpretable directed graphs. A dual-branch architecture further integrates whole-
brain connectivity with region-specific processing aligned to established functional neuroanatomy. On the MEEG
dataset, GL-LFGNN achieves 86.17% (Arousal) and 86.71% (Valence) accuracy with only 37K parameters – approximately 10% of the current state-of-the-art – demonstrating that principled causal modeling can simultaneously enhance interpretability, generalization, and computational efficiency. Code will be released.
Read the full abstract ▾
Machine Learning · Cross-listed from math.NA
Random Neural Network Expressivity for Non-Linear Partial Differential Equations
Muhammed Ali Mehmood and Lukas Gonon
Neural networks with randomly generated hidden weights (RaNNs) have been extensively studied, both as a standalone learning method and as an initialization for fully trainable deep learning methods. In this work, we study RaNN expressivity for learning solutions to non-linear partial differential equations (PDEs). Despite their widespread use in practical applications, a rigorous theoretical understanding of the approximation properties of RaNNs in this context remains limited. Here, we derive error bounds for RaNN approximations to time-dependent Sobolev functions and obtain a dimension-free approximation rate $\frac{1}{2}$ for sufficiently regular functions. We apply our results to two important classes of non-linear PDEs: Porous Medium Equations and Compressible Navier-Stokes Equations, showing that RaNNs are capable of efficiently approximating solutions to these complex, non-linear PDEs. Our theoretical analysis is supported by numerical experiments, showing that the obtained convergence rates extend beyond the considered setting.
Read the full abstract ▾
Machine Learning
Scale When Needed: Adaptive Neuron-level Mixed Precision Quantization Aware Training
Ayush K. Varshney, Konstantinos Vandikas, Šarūnas Girdzijauskas et al. (+2 more)
Deploying deep neural networks on resource-constrained 6G edge devices demands aggressive compression with minimal accuracy loss. Quantization-Aware Training (QAT) has emerged as a leading compression approach; however, existing mixed-precision methods typically operate at coarse layer- or channel-level granularity. These methods often rely on heuristic or search-based bit-allocation strategies, which may overlook fine-grained variability at the
neuron level. We propose
Neuron-Level Mixed-Precision QAT (NMP-QAT), where each
neuron independently learns its own discrete precision during training. Starting from low-bit precision, NMP-QAT expands bit-width only when training signals demand it, via differentiable surrogates and straight-through
estimators, while preserving a fully discrete inference graph. This adaptability extends to both weights and activations, reducing
memory movement. Evaluated on telecom and non-telecom
datasets across MLP and tabular foundation model architectures, NMP-QAT achieves superior compression-accuracy trade-offs over mixed-precision QAT baselines, making it well-suited for Green AI deployments at the network edge.
Read the full abstract ▾
All authors Ayush K. Varshney, Konstantinos Vandikas, Šarūnas Girdzijauskas, Adam Orucu, Aneta Vulgarakis Feljan
Machine Learning · Cross-listed from stat.AP
Multimodality Stacking with Blockwise missing values and application to the PIONeeR biomarkers study for prediction of resistance to immunotherapy
Mohamed Boussena, Florence Monville, Jacques Fieschi-Meric et al. (+8 more)
Integrating
multimodal datasets in clinical oncology is frequently hindered by high dimensionality and blockwise missingness, where entire data sources are unavailable for specific patient subsets. Standard survival models often struggle with these gaps, leading to biased results or patient exclusion. We introduce Multimodality Stacking with Blockwise missing values (MSB), a late-fusion framework for survival analysis that independently models modality-specific features before aggregating predictions via a cross-validated stacking meta-learner. MSB was validated on the PIONeeR study (n=443 patients, 378 biomarkers across eight heterogeneous sources) to predict progression-free survival in advanced non-small
cell lung cancer patients receiving immunotherapy. MSB yielded higher predictive performance (C-index) than baseline
algorithms. Improvements varied by baseline strength: linear models showed a 15.9% increase (p<0.001 for the Wilcoxon signed-rank test), random survival forests gained 5.4% (p=0.002), and gradient boosting methods improved by 2.1% (p=0.030). Beyond discrimination, MSB reduced the generalization gap (train-test difference in 5 folds cross-validation repeated 3 times: 0.055 vs 0.380 for linear models). Permutation importance analysis identified routine laboratory markers, clinical features, and PD-L1 expression as primary predictive drivers. Missing block indicators showed negligible importance, suggesting the model learned from biomarker values rather than data availability patterns. MSB provides a statistically validated framework for
multimodal survival prediction with blockwise missingness. By enabling systematic biomarker evaluation without requiring complete data, MSB offers a practical tool for predictive modeling in biomedical research, pending external validation. Implementation is available at https://github.com/MohamedBoussena/MSB under Inria license.
Read the full abstract ▾
All authors Mohamed Boussena, Florence Monville, Jacques Fieschi-Meric, Frederic Vely, Pierre Milpied, Julien Mazieres, Maurice Perol, Eric Vivier, Laurent Greillier, Fabrice Barlesi, Sebastien Benzekry
Machine Learning · Cross-listed from cs.CL
TRACE: A taxonomy-grounded synthetic dataset for teaching-program generation and session interpretation in Applied Behavior Analysis
Festus Kahunla
Applied Behavior Analysis (ABA) is a clinical discipline whose documentation, teaching programs and multi-session behavioral logs, is formulaic and high-volume, yet real session data is HIPAA-protected and bound by professional confidentiality rules, blocking the release of a training corpus. We present TRACE (Taxonomy-Referenced ABA Clinical Examples), a 2,999-example synthetic instruction-tuning
dataset covering two ABA tasks: teaching-program generation across Discrete Trial Training, Natural Environment Teaching, and Task Analysis; and multi-session behavioral interpretation across twelve trajectory patterns and thirteen target behaviors. Every example is produced by a deterministic taxonomy-driven generator grounded in the canonical ABA literature, and every example carries complete sampling provenance, the exact taxonomy
cells that produced it. The
dataset is released under CC BY-NC 4.0 for data and MIT for code, with stratified train (2,549), validation (149), test (281), and sanity (20) splits. TRACE is a research artifact and has not been clinically validated.
Read the full abstract ▾
Machine Learning
MimirRAG: A Multi-Agent RAG Framework for Financial Data Retrieval with Metadata Integration
Magnus Samuelsen, Wilmer Nyström, Somnath Mazumdar et al. (+2 more)
Retrieval-augmented generation (RAG) systems offer a promising approach to reduce
hallucinations and improve answer accuracy in large
language models (LLMs), a requirement for reliable, financial analysis where answers must be grounded in verifiable evidence from filings rather than generated from model priors. However, designing RAG systems that extract meaningful insights from mixed financial documents and integrate into analyst workflows remains challenging. This paper introduces MimirRAG (Metadata-Integrated Multi-
Agent Information
Retrieval), a multi-
agent RAG system developed iteratively to address these challenges. MimirRAG features a modular pipeline encompassing structure-preserving parsing of PDF filings, table-aware chunking, metadata extraction,
agent-based
retrieval with query planning and hybrid search, validation, and context-aware generation with numerical
reasoning support. Our ablation study identifies three key technical enablers for effective financial RAG: metadata integration, table-aware chunking, and an agentic workflow. MimirRAG was evaluated quantitatively using FinanceBench and qualitatively through expert validation with four financial analysts. The system achieved 89.3% accuracy on FinanceBench, outperforming the original
benchmark baselines. Expert feedback highlighted that successful deployment also requires calibrated trust, comprehensive data integration, and user personalization. We conclude that combining multi-
agent RAG architecture with human-centric design principles can improve the extraction of meaningful insights in financial analysis.
Read the full abstract ▾
All authors Magnus Samuelsen, Wilmer Nyström, Somnath Mazumdar, Mansoor Hussain, Mikkel Strange
Machine Learning
A perspective on fluid mechanical environments for challenges in reinforcement learning
Shruti Mishra, Michael Chang, Vamsi Spandan et al. (+1 more)
We consider the challenge of developing
agents that efficiently interact with high-dimensional, evolving environments, towards a view of practical reinforcement learning (RL)
agents interacting with open worlds, of which they witness and affect only a small part. We argue that canonical fluid mechanics problems, and their simulations, present a compelling testbed for the development of such methods. These problems arise in nonlinear instabilities, where small disturbances can grow to transform the dynamics of a system. Nonlinear instabilities represent several open scientific challenges with industrial applications – the droplet breakup of a
liquid jet, mixing at an interface between two fluids, and the appearance of unusually tall rogue waves in the ocean. In these settings,
agents may leverage preserved representations across the changing dynamics to learn efficiently. We present two problem descriptions of
agents interacting with a fluid mechanical environment, and describe the state and action spaces, and reward functions, for these
agents. For these examples, we specify the aspects of the environment which are nonstationary and the preserved invariances. We note Dedalus and JAX-CFD as open-source simulators that can be used for the development of reinforcement learning methods (Burns et al., 2016; Kochkov et al., 2021)) We demonstrate the use of Dedalus for environment generation by creating RL
agents that learn to navigate in a stationary environment that is simulated using Dedalus. This sets the stage for future development of RL
agents that learn to meaningfully interact with simulated environments that represent scientific challenges in natural and industrial flows.
Read the full abstract ▾
All authors Shruti Mishra, Michael Chang, Vamsi Spandan, Shmuel M. Rubinstein
Machine Learning · Cross-listed from cs.RO
Convex-Neural RRT*: Fast and Reliable Learning-Guided Sampling for High-Quality Robot Path Planning
Hichem Cheriet, Badra Khellat Kihel, Samira Chouraqui et al. (+1 more)
Sampling-based
algorithms for
robot path planning offer probabilistic completeness and strong empirical convergence properties across environments with diverse obstacle configurations. However, in practice, these methods often require many iterations to obtain high-quality solutions. This paper proposes Convex-Neural RRT*, an enhanced RRT* variant that incorporates neural guidance to predict informative waypoint regions near high-quality paths. Convex candidate regions are extracted from these predictions, enabling the planner to concentrate exploration on geometrically relevant areas while preserving global exploration. The proposed
algorithm is evaluated against Neural RRT*, Neural Informed RRT*, classical RRT*, and LTA* across three environment types and 18
benchmark maps. Experimental results show that Convex-Neural RRT* reduces computation time by 30-75% compared to neural-guided variants and up to 88-98% relative to LTA*, while achieving an average path length reduction of approximately 5% compared to classical RRT*, with larger improvements observed in complex environments. The method also maintains an overall success rate above 99% across varying obstacle densities. These findings indicate that convex-guided neural sampling provides an effective balance between computational efficiency and solution quality, supporting its applicability to time-sensitive robotic
navigation tasks.
Read the full abstract ▾
All authors Hichem Cheriet, Badra Khellat Kihel, Samira Chouraqui, Bara J. Emran
Machine Learning
Metropolis-Scale Resilient and Trustworthy Traffic Flow Inference Using Multi-Source Data
Qishen Zhou, Yifan Zhang, Michail A. Makridis et al. (+3 more)
Inferring network-wide traffic states from sparse observations with high accuracy and trustworthy uncertainty quantification is essential for intelligent transportation systems, yet it remains challenging due to the underdetermined nature of the problem, multifaceted disturbances in sensing networks, and the inherent conflicts among multiple inference subtasks when modeled jointly. We propose the Task-Aware Attentive Neural Process (TA-ANP), a unified probabilistic framework for resilient and trustworthy global traffic state inference (GTSI) by fusing floating car data (FCD) with sparse fixed-detector measurements. By casting GTSI as a
stochastic process, TA-ANP leverages the meta-learning properties of neural processes to adapt rapidly to changes in sensing configurations without retraining. A task-aware multi-query
attention module with distinct spatiotemporal inductive biases is introduced to handle three GTSI subtasks jointly —real-time estimation at unobserved locations,
forecasting at both observed and unobserved locations—while mitigating cross-task interference. For uncertainty quantification, we combine neural processes with Monte Carlo Dropout to capture both aleatoric and epistemic uncertainty. To support metropolis-scale evaluation, we construct the Metropolitan Multi-Source Traffic
Dataset (MMTD), integrating fixed-loop sensor measurements, FCD statistics, and OpenStreetMap road-network data over an urban network of 26,166 road segments, including 2,371 fixed-sensor-equipped segments used for quantitative evaluation. Experiments across multiple independently generated sensor configurations show that TA-ANP achieves best performance across all subtasks under deterministic and probabilistic metrics. The resulting well-calibrated uncertainties enable more efficient fixed-sensor placement with fewer sensor deployments.
Read the full abstract ▾
All authors Qishen Zhou, Yifan Zhang, Michail A. Makridis, Anastasios Kouvelas, Yibing Wang, Simon Hu
Machine Learning
Mitigating Gradient Pathology in PINNs through Aligned Constraint
Yichen Luo, Peiyu Zhu, Dongxiao Hu et al. (+5 more)
While Physics-Informed Neural Networks (PINNs) are powerful for solving Partial Differential Equations (PDEs), their training is often paralyzed by gradient pathology. The gradients from the PDE residuals and boundary constraints oppose each other, trapping the model in local minima. Current solutions, such as adaptive weighting or hard constraints, either fail to fundamentally resolve this ill-conditioning or are limited to simple geometries. In this study, we systematically analyze the possible causes of this gradient pathology from the perspectives of loss landscapes and optimization dynamics. Based on the obtained conclusion, we propose Constraint-Aligned loss with
Manifold Lifting (CAML). By reformulating all zeroth-order terms into aligned constraints, our method effectively mitigates gradient conflicts. In addition, we introduce a delay factor to help the optimizer skip the high-
curvature area. Experiments demonstrate that our CAML significantly enhances numerical stability and efficiency in highly complex PINN problems. Our code is open-sourced on https://github.com/YichenLuo-0/CAML.
Read the full abstract ▾
All authors Yichen Luo, Peiyu Zhu, Dongxiao Hu, Jia Wang, Tailin Wu, Dapeng Lan, Yu Liu, Zhibo Pang
Machine Learning · Cross-listed from cs.NI
Scaling up Energy-Aware Multi-Agent Reinforcement Learning for Mission-Oriented Drone Networks with Individual Reward
Changling Li and Ying Li
Multi-
agent reinforcement learning (MARL) has shown wide applicability in collaborative systems such as autonomous driving and smart cities for its ability of learning through interaction. With the recent development of drone networks, researchers have also applied MARL to address the trajectory planning problems. However, the dynamic environment and the limited battery capacity are still challenging for using MARL to achieve efficient collaborative task execution. In this paper, we propose an energy-aware MARL model as an attempt to tackle these challenges, leveraging Deep Q-Networks (DQN) with individual reward functions driven by the task execution progress and the remaining battery of drones. We conduct a set of simulation studies for the proposed mode and compare it with the shared reward MARL~\cite{Li2022MARL} to explore the impact of credit assignment in MARL. The results indicate that our proposed model can achieve at least 80% success rate regardless of the task locations and lengths. Similar to the shared reward mode, the individual reward mode can achieve a better success rate when the task density is high, and it can hit nearly a 100% success rate when task density gets close to 40%. The true advantage of our proposed model with individual reward is revealed when scaling up the environment. The comparison to the shared reward MARL shows that the our proposed model is more robust towards the change of the environment size and
agent numbers. It can achieve higher success rate with fewer steps due to the clarity of the goal which improves energy efficiency even better.
Read the full abstract ▾
Machine Learning
Selective Test-Time Compute Scaling for Click-Through Rate Prediction via Uncertainty-Triggered Feature Path Exploration
Moyu Zhang, Yun Chen, Yujun Jin et al. (+3 more)
Scaling test-time compute has proven highly effective for
language models, yet this opportunity remains largely unexplored for industrial Click-Through Rate (CTR) prediction. CTR models suffer from a fundamental asymmetry: feature combinations well-represented in training yield confident predictions, while sparsely observed ones produce unreliable outputs. Existing training-phase solutions such as adaptive gating learn a fixed selection function subject to the same sparsity, offering no per-instance recourse at deployment.We propose UTTSI (Uncertainty-Triggered Test-Time Selective Inference), a training-free model-agnostic framework that scales inference depth proportionally to per-instance uncertainty. A dual-signal
estimator combining model logit confidence with a data-level frequency prior distinguishes epistemic uncertainty from aleatoric ambiguity. Every instance undergoes adaptive feature filtering to remove unreliable
embeddings; uncertain instances additionally receive
stochastic feature-path explorations whose predictions are aggregated via consistency-weighted ensembling. Confident instances bypass exploration entirely, keeping average overhead at approximately $2.8\times$ base model cost with worst-case latency unchanged.Experiments on four
datasets with three backbone architectures demonstrate consistent, statistically significant gains over all training-phase baselines. A seven-day online A/B test further confirms a 5.3% relative CTR gain ($p < 0.01$), establishing selective test-time compute allocation as a practical complement to training-phase advances for CTR prediction.
Read the full abstract ▾
All authors Moyu Zhang, Yun Chen, Yujun Jin, Jinxin Hu, Yu Zhang, Xiaoyi Zeng
Machine Learning · Cross-listed from cs.IR
Self-Balancing Gradient Allocation for Heterogeneity-Aware Feature Generation in Click-Through Rate Prediction
Moyu Zhang, Yun Chen, Yujun Jin et al. (+3 more)
Generative pre-training via discrete
diffusion provides dense reconstruction supervision across all feature fields simultaneously, mitigating representation collapse from data sparsity in CTR prediction. However, all existing generative CTR methods share a fundamental limitation: the reconstruction objective assigns equal training weight to every feature field, ignoring the profound heterogeneity of reconstruction difficulty across high-cardinality ID fields, sparse categorical attributes, numerical values, and behavioral sequences. This causes easy fields to dominate training gradients while the hardest but most informative fields remain chronically underfit, a problem we term the generative difficulty imbalance.We propose HeteGenCTR, which resolves this imbalance through per-field learnable difficulty parameters jointly trained with the denoising network. This unified signal drives two coordinated components without additional hyperparameters: a self-balancing loss that automatically reallocates gradient budget toward harder fields with a provably stable
equilibrium, and a difficulty-guided
attention mechanism that suppresses the influence of already-converged easy fields while amplifying cross-field information flow toward hard fields. Both components share the same learned signal and remain mutually consistent throughout training. Experiments on five CTR
benchmarks and a seven-day online A/B test demonstrate consistent, statistically significant improvements over state-of-the-art baselines, with disproportionate gains for cold-start and long-tail users.
Read the full abstract ▾
All authors Moyu Zhang, Yun Chen, Yujun Jin, Jinxin Hu, Yu Zhang, Xiaoyi Zeng
Machine Learning · Cross-listed from cs.RO
Learning, locomotion, and navigation of soft synthetic snakes in three-dimensional, heterogeneous environments
Xiaotian Zhang, Ali Albazroun, Tixian Wang et al. (+3 more)
Limbless terrestrial animals exhibit exceptional locomotor versatility and control, currently unmatched by engineered counterparts. Here, we introduce a computational framework that enables soft synthetic snakes to navigate unstructured, heterogeneous 3D terrains. Our approach is grounded in bio-inspired actuation and sensing models that reduce the control complexity inherent to high-degree-of-freedom, continuum bodies. These models are integrated into a reinforcement learning architecture to derive environment-traversing policies. Training first occurs in simplified, homogeneous terrains to learn locomotion primitives. These are then composed into adaptive strategies for complex landscapes. We demonstrate robustness by deploying a snake in high-fidelity 3D environments reconstructed from real-world imaging, achieving reliable
navigation. Overall, this work provides a physically-realistic simulation platform and practical insights for the control of continuum systems in natural terrains.
Read the full abstract ▾
All authors Xiaotian Zhang, Ali Albazroun, Tixian Wang, Songyuan Cui, Prashant G. Mehta, Mattia Gazzola
Machine Learning
Benchmarking non-conformity score functions in conformal prediction
Sol Erika Boman
Conformal prediction is a useful and versatile alternative to model calibration in machine learning classification. It replaces single-class prediction with prediction sets, guaranteeing that the a priori probability of the prediction sets containing the true class is larger than or equal to a pre-specified rate. The size and usefulness of the prediction sets relies heavily on the choice of the non-conformity score function. The scientific literature contains many examples of non-conformity score functions but there is an absence of studies examining their properties and effectiveness. In this paper, we give an overview of properties of non-conformity score functions. We give examples of non-conformity score functions in the existing literature and introduce original modifications. We introduce an original method of evaluating the prediction set sizes of conformal predictors and use it to provide a comparison between non-conformity score functions. We also examine efficacy of different non-conformity score functions for class-conditional conformal prediction in a setting with imbalanced classes.
Read the full abstract ▾
Machine Learning · Cross-listed from cs.CL
Large Language Model Selection with Limited Annotations
Yavuz Durmazkeser, Patrik Okanovic, Andreas Kirsch et al. (+2 more)
Choosing a Large
Language Model (LLM) for a given task requires comparing many strong candidates, yet standard evaluation relies on costly annotations over fixed evaluation sets. To address this challenge, we develop SELECT-LLM, the first framework for active model selection of LLMs. SELECT-LLM aims to find a small set of queries whose annotations are most informative for identifying the best LLM for a given task. To this end, we introduce a query selection rule based on expected information gain, computed from pairwise similarities between candidate model outputs. Because this rule only uses generated model responses, SELECT-LLM can be applied across candidate models without assumptions about their architecture or access to model weights. This makes it suitable for both open-weight and black-box LLMs. We evaluate SELECT-LLM across 23
datasets, 156 evaluated models, diverse task families, and multiple text evaluation metrics. Across all experiments, SELECT-LLM improves over the strongest baseline in every setting, with annotation cost reductions up to 81.8% for best model selection and up to 84.78% for near-best model selection.
Read the full abstract ▾
All authors Yavuz Durmazkeser, Patrik Okanovic, Andreas Kirsch, Torsten Hoefler, Nezihe Merve Gürel
Machine Learning · Cross-listed from cs.RO
Bridging the Gap: Enabling Soft Actor Critic for High Performance Legged Locomotion
Gianluca Sabatini, Chenhao Li and Marco Hutter
Proximal Policy Optimization (PPO) has become the de facto standard for training legged
robots, thanks to its robustness and scalability in massively parallel simulation environments like IsaacLab. However, its on-policy nature makes it inherently sample-inefficient, preventing its use for continuous adaptation and fine-tuning on real hardware. Soft Actor-Critic (SAC), by contrast, is an off-policy
algorithm that can reuse past experience, making it a natural candidate for sim-to-real transfer workflows where the same
algorithm can be used both in simulation and for online learning on the real
robot. Despite these advantages, SAC has consistently failed to match PPO's empirical performance in massively parallel training settings. This work identifies the root causes of this gap and introduces targeted modifications, covering policy initialization, timeout-aware critic targets, and multi-step return estimation, that enable SAC to train stably at scale. Evaluated across multiple legged
robot platforms and diverse locomotion tasks, our approach closes the performance gap with PPO entirely.
Read the full abstract ▾
Machine Learning
TGFormer: Towards Temporal Graph Transformer with Auto-Correlation Mechanism
Hongjiang Chen, Pengfei Jiao, Ming Du et al. (+4 more)
The growing interest in Temporal Graph Neural Networks (TGNNs) stems from their ability to model complex dynamics and deliver superior performance. However, TGNNs encounter fundamental challenges in capturing long-term dependencies and identifying periodic patterns. To address these limitations, we propose TGFormer, a novel
Transformer architecture specifically designed for temporal graphs. Our model redefines temporal graph learning by establishing a trajectory framework that aligns with time series analysis principles. This approach allows TGFormer to derive node representations through systematic analysis of historical interactions, enabling granular examination of node relationships across sequential timestamps. Building upon
stochastic process theory, we develop an auto-correlation mechanism that systematically uncovers periodic dependencies in node interactions. This innovation empowers TGFormer to perform dependency discovery and representation aggregation at sub-interaction levels, demonstrating superior efficiency and accuracy compared to conventional
attention mechanisms. Experimental validation across six public
benchmarks confirms the effectiveness of our approach, with TGFormer at most achieving 9.35% precision improvement compared to state-of-the-art approaches.
Read the full abstract ▾
All authors Hongjiang Chen, Pengfei Jiao, Ming Du, Xuan Guo, Zhidong Zhao, Di Jin, Xiao Liu
Machine Learning
OSDTW: Optimal Shared Depth and Task Weighting for Long-Tailed Recognition
Chang Chu, Qingyue Zhang, Shao-Lun Huang et al. (+1 more)
Long-tailed recognition suffers from a persistent head–tail trade-off: improving tail performance often degrades head accuracy and can increase training instability. Despite strong empirical results from re-weighting, decoupled training, and multi-expert methods, key design choices about representation sharing between head and tail classes and supervision weighting across class groups remain largely heuristic. In this work, we propose OSDTW, a principled task-decomposition framework that partitions the original single-label recognition problem into a head task and a tail task, implemented with a shared encoder and task-specific decoders. To handle the mutual exclusivity and statistical dependence between the two label groups, we introduce a factorized model and show that the resulting Kullback–Leibler divergence-based generalization error can be written as the sum of task-wise terms up to an additive constant, yielding a well-defined task-wise objective. We further develop a three-stage training pipeline: independent task training to estimate task-wise optima and the Fisher information
matrix, weighted joint training to learn a shared encoder, and branch assembly to construct the final decoupled model. Under a block-diagonal Fisher approximation, we derive a computable second-order expansion of the expected generalization error, decomposing it into encoder variance, encoder bias, and decoder variance. This bias–variance decomposition provides a computable proxy to select the shared depth and task weights, enabling efficient hyper-parameter search. Experiments on standard long-tailed
benchmarks demonstrate the effectiveness of the proposed approach over strong baselines.
Read the full abstract ▾
All authors Chang Chu, Qingyue Zhang, Shao-Lun Huang, Junxiong Zheng
Machine Learning · Cross-listed from cs.CV
Cross-Domain Generalization Limits of Vision Foundation Models in Facial Deepfake Detection
Ibrahim Delibasoglu
The rapid
evolution of generative models has enabled the creation of hyper-realistic facial deepfakes, exposing a critical vulnerability in modern digital forensics: the inability of detectors to generalize to unseen manipulation techniques. Traditional networks suffer from representation collapse, overfitting to localized artifact fingerprints of specific training generators. This work investigates whether modern
Vision Foundation Models can serve as generalizable, out-of-the-box feature extractors capable of tracking forensic anomalies across entirely unseen generative
manifolds. We conduct a systematic cross-domain evaluation comparing three foundational learning paradigms: fully supervised macro-semantic features (RoPE-ViT), pure self-supervised
geometric features (DINOv3), and multi-teacher agglomerative representations (NVIDIA C-RADIOv4-H). By deploying frozen backbones subjected to downstream linear probing, we map the performance limitations of these architectures on the challenging DF40
benchmark. Our empirical findings expose the intrinsic trade-offs between pre-training paradigms and parameter scale, proving that while foundation models retain high discriminative capabilities for entire face synthesis, localized face editing techniques expose fundamental boundaries in linear probe evaluation structures. Source code and model weights are available in http://github.com/mribrahim/deepfake
Read the full abstract ▾
Machine Learning
MedMamba: Multi-View State Space Models with Adaptive Graph Learning for Medical Time Series Classification
Da Zhang, Bingyu Li, Zhiyuan Zhao et al. (+3 more)
Medical time series are central to healthcare, enabling continuous monitoring and supporting timely clinical decisions. Despite recent progress, existing methods struggle to jointly model local-global dynamics and handle nonstationarities like baseline drift, while often failing to capture latent channel interactions. To address these challenges, we propose MedMamba, an end-to-end architecture that integrates state space models with domain-specific inductive biases. Specifically, MedMamba first employs multi-scale convolutional
embeddings to capture discriminative local morphology. Second, to mitigate nonstationarity, we introduce a tri-branch differential state space encoder that processes raw, temporal-difference, and frequency-domain views, fusing them to emphasize informative patterns while suppressing drift. Furthermore, to uncover latent channel correlations, we design a spatial graph Mamba module that learns a directed dependency structure regularized toward sparsity and acyclicity, which obviates the need for predefined graphs. Extensive experiments on five real-world
datasets demonstrate that MedMamba achieves state-of-the-art performance while maintaining linear computational complexity, and ablation studies validate each component's contribution.Code is available at https://github.com/zhangda1018/MedMamba.
Read the full abstract ▾
All authors Da Zhang, Bingyu Li, Zhiyuan Zhao, Hongyuan Zhang, Junyu Gao, Xuelong Li
Machine Learning · Cross-listed from cs.CL
Investigating the Interplay between Contextual and Parametric Chain-of-Thought Faithfulness under Optimization
Jingyi Sun, Qianli Wang, Pepa Atanasova et al. (+2 more)
Chain-of-Thought (CoT) faithfulness, i.e., whether CoTs genuinely reflect large
language models' (LLM) underlying behavior, is typically evaluated with metrics under two disjoint paradigms: contextual faithfulness, measured by perturbing the input or CoT trace, and parametric faithfulness, assessed by intervening on a model's parametric knowledge. Yet prior work compares them only descriptively. We fill this gap by proposing FaithMATE, a unified preference-
alignment interface for optimizing models towards either faithfulness paradigm. It enables us to investigate the interplay between the two paradigms, examining whether and to what extent faithfulness gains generalize within and across paradigms. Across three models, two
datasets, and six faithfulness metrics, we find that the two paradigms are positively coupled, yet asymmetric: optimizing towards parametric faithfulness yields consistent gains across both paradigms, whereas the contextual counterpart delivers more variable gains. Within the contextual paradigm, faithfulness gains on one metric do not consistently transfer to others, implying that existing contextual metrics capture disjoint facets of faithfulness and exposing inherent trade-offs. These findings imply that CoT faithfulness is not a monolithic objective and therefore requires multifaceted optimization and evaluation.
Read the full abstract ▾
All authors Jingyi Sun, Qianli Wang, Pepa Atanasova, Nils Feldhus, Isabelle Augenstein
Machine Learning · Cross-listed from cs.AI
Yuanzhi Xu, Qian Gao, Jun Fan et al. (+4 more)
The generation of factually incorrect objects, commonly known as object
hallucination, remains a persistent challenge in Large
Vision-
Language Models (LVLMs). Current approaches to address this issue - ranging from expensive data-driven fine-tuning and high-latency contrastive decoding to rigid
attention head truncation - frequently compromise either computational efficiency or the continuity of the model's feature space. To overcome these limitations, we introduce a novel, training-free inference strategy that operates as a region-aware adaptive weighting mechanism to dynamically correct semantic drift without relying on abrupt heuristic truncations. By computing an outlier-resistant statistical midpoint across various
attention heads, we establish a stable anchor for reliable visual representations. We then utilize the inter-head disagreement mapped across regions to dynamically determine intervention budgets, gently suppressing
hallucination-inducing
attention paths through a continuous penalty modulation. This recalibration process effectively rectifies visual-semantic misalignments while fully preserving generative fluency and
language priors. Comprehensive evaluations on standard
multimodal benchmarks, including CHAIR, POPE, and MME, reveal that our strategy substantially curtails both instance- and sentence-level
hallucinations. The results demonstrate state-of-the-art performance against contemporary baselines, confirming our method's efficiency and algorithmic robustness. Our code will be public.
Read the full abstract ▾
All authors Yuanzhi Xu, Qian Gao, Jun Fan, Guohui Ding, Zhenyu Yang, Sixue Lin, Yuteng Xiao
Machine Learning · Cross-listed from cs.RO
ARCANE-PedSynth: Synthetic Multi-Pedestrian Datasets with Behavioural Crossing Annotations
Muhammad Naveed Riaz, Maciej Wielgosz and Antonio M. López Peña
We present ARCANE-PedSynth, an open-source CARLA-based software framework for generating synthetic multi-pedestrian
datasets with dense behavioural annotations for pedestrian crossing prediction in autonomous driving. The framework overcomes CARLA's native 9% crossing rate through a hybrid AI-manual pedestrian control architecture, enabling configurable target rates up to 75%. A 12-state behavioural finite state machine with five character archetypes produces diverse crossing behaviours. The framework generates synchronised RGB,
LiDAR, and DVS data with per-frame crossing labels, behavioural states, and estimated 2D pose keypoints. We demonstrate ARCANE-PedSynth through PedSynth++, an example
dataset generated with the framework, comprising 533 multi-pedestrian clips across 12 weather conditions with RGB,
LiDAR, and DVS streams. ARCANE-PedSynth is fully reproducible via CLI parameterisation and Docker containerisation.
Read the full abstract ▾
Machine Learning
RealBench: Benchmarking Data-Driven Numerical Weather Forecasting Under Operational Conditions and Extreme Event Challenges
Ruize Li, Zhibin Wen, Tao Han et al. (+5 more)
Accurate evaluation of weather
forecasting models is critical for their reliable deployment in real-world applications. However, existing
benchmarks predominantly rely on reanalysis products such as ERA5, which are generated through delayed data assimilation and do not reflect the constraints of real-time operational
forecasting, thereby resulting in a systematic mismatch between
benchmark performance and real-world
forecasting. In this work, we introduce RealBench, a next-generation
benchmark for AI weather
forecasting that emphasizes realistic evaluation under operational conditions. RealBench features a strictly out-of-distribution test set spanning 2025 to eliminate data leakage and capture recent atmospheric regimes. It integrates multiple data sources, including low-latency operational analysis and a large-scale global in-situ observation
dataset comprising over 10,000 stations, enabling direct evaluation against real atmospheric measurements. Beyond standard global metrics, RealBench provides a comprehensive evaluation framework for high-impact extreme events, including heatwaves, cold surges, and tropical cyclones, using event-specific metrics that better reflect real-world
forecasting priorities. The evaluation results reveal substantial discrepancies between reanalysis-based metrics and real-world performance, particularly concerning extreme events. By highlighting the limitations of existing
benchmarks, this work establishes a more faithful and operationally relevant evaluation paradigm, providing a rigorous foundation for advancing next-generation AI weather
forecasting systems. The
benchmark implementation is available at: https://github.com/lixruize-del/NWP-
Benchmark.
Read the full abstract ▾
All authors Ruize Li, Zhibin Wen, Tao Han, Hao Chen, Fenghua Ling, Wei Zhang, Song Guo, Lei Bai
Machine Learning
Riemannian-Manifold Steering: Geometry-Aware Generative Autoencoders for Label-Free Steering
Narmeen Oozeer, Shivam Raval, Philip Quirke et al. (+4 more)
Steering a
language model - intervening on its internal activations to change downstream behaviour - has recently expanded beyond linear interpolation to nonlinear methods such as angular and kernelized steering, which define intervention transformations without learning an explicit geometry over paths in activation space. Freshly introduced geometry-aware
manifold methods do learn such a geometry, but require labelled class centroids together with prescribed cyclic or sequential structure. These assumptions restrict where
manifold steering can be applied, since existing constructions require labelled centroids and compatible boundary conditions. We recast
manifold steering more broadly as Riemannian geodesic computation on activation space, recovering linear and labelled-spline steering as geodesics under particular choices of metric. A principled metric within this framework is the output-space Hellinger distance pulled back to activations; we approximate this with a learned encoder trained on output distances over a small concept-
token schema - no per-
prompt labels, no
topology prior, and no per-task curve fitting. Empirically, the method reliably drives the model onto the target class across all tasks in a standard four-task
language-model arithmetic
benchmark, while following more behaviourally natural trajectories than baselines on smaller output spaces. We thereby provide a unified Riemannian framework for
manifold steering together with a schema-supervised, label-free instantiation that operates without labelled centroids or prescribed boundary conditions.
Read the full abstract ▾
All authors Narmeen Oozeer, Shivam Raval, Philip Quirke, Manikandan Ravikiran, Jeff Phillips, Shriyash Upadhyay, Amirali Abdullah
Machine Learning · Cross-listed from cs.CR
Memory-Induced Tool-Drift in LLM Agents
Mahavir Dabas, Jihyun Jeong, Ming Jin et al. (+1 more)
LLM
agents increasingly combine long-term
memory for personalization with tool calling for taking actions. We study a failure of this combination: personality-driven preferences stored in
memory (e.g., cost-consciousness, impatience, risk tolerance) silently shifting tool-call parameters in contexts where they do not apply. We call this
memory-induced tool-drift and stress-test it with MEMDRIFT, an iteratively refined
benchmark of 105 scenarios spanning five bias dimensions and seven professional domains. Measured as excess parameter degradation over a neutral-
memory baseline, biased
memories in MEMDRIFT induce drift in all eight models we evaluate, raising degradation rates to 0.38–0.83 versus 0.05–0.12 for neutral
memories. The drift persists under three widely used
memory frameworks. A vulnerability audit of 6,062 tools across 288 MCP servers demonstrates that real-world tools are also susceptible to this failure mode. Mechanistically, we find that biased
memories act as implicit steering vectors along the same direction as explicit behavioral instructions and draw
attention away from task-relevant context. Neither
prompt-based defenses nor relevance filters offer a complete solution, motivating defenses at the intersection of
memory management and tool calling.
Read the full abstract ▾
All authors Mahavir Dabas, Jihyun Jeong, Ming Jin, Ruoxi Jia
Machine Learning
Global linear convergence of entropy-regularized softmax policy gradient beyond tabular MDPs
Ziyue Chen, David Šiška and Lukasz Szpruch
We study the global convergence of policy gradient for infinite-horizon
entropy-regularized Markov decision processes (MDPs) with continuous state and action spaces. We consider log-linear softmax policies with linear function approximation, which extend the tabular softmax parameterization while retaining a tractable policy class. Under $Q^\pi_\tau$-realizability for the regularized state-action value function, we first establish a non-uniform Polyak–Łojasiewicz (PŁ) inequality. The non-uniformity arises through degeneracy of constants associated with the policy geometry, namely the Fisher information
matrix or an uncentered feature covariance
matrix. We then identify two feature regimes under which this non-uniform constant can be bounded along the gradient flow. For full-affine-span features, we prove radial unboundedness of the KL regularizer and show that the smallest
eigenvalue of the Fisher information
matrix remains bounded below by an initialization-dependent positive constant. For simplex-valued features, we prove an analogous radial unboundedness result in the subspace orthogonal to the all-ones vector and obtain a uniform lower bound for the smallest
eigenvalue of the uncentered covariance
matrix. These results imply global linear convergence of the regularized objective along the gradient flow, i.e. suboptimality decaying as $\mathcal{O}(e^{-Ct})$ for some $C>0$. Our analysis extends the global convergence theory of
entropy-regularized softmax policy gradient beyond the tabular setting of Agarwal et al. (2020); Bhandari and Russo (2024); Mei et al. (2020).
Read the full abstract ▾
Machine Learning · Cross-listed from cs.RO
HumanEgo: Zero-Shot Robot Learning from Minutes of Human Egocentric Videos
Zhi Wang, Botao He, Kelin Yu et al. (+4 more)
Human
egocentric video captures rich manipulation demonstrations without any
robot hardware, yet transferring these skills to
robots remains challenging due to the embodiment gap between human and
robot in both visual appearance and kinematics. We present HumanEgo, a framework that bridges the embodiment gap by lifting each human demonstration to an entity-level representation of hand-object interaction, and training a flow matching policy with dense auxiliary objectives that amplify supervision from every trajectory. HumanEgo is
robot-data-free, hardware-agnostic, data-efficient, and zero-shot human-to-
robot transferable. With only 30 minutes of human
videos per task, HumanEgo achieves 92.5% average success across four real-world tasks (75% with just 15 minutes), outperforms matched-time
robot teleoperation by 41%, and robustly transfers zero-shot across novel
robots, cameras, and environments. We release HumanEgo as an easy-to-use, open-source framework for learning
robot policies directly from human data: https://github.com/TX-Leo/HumanEgo
Read the full abstract ▾
All authors Zhi Wang, Botao He, Kelin Yu, Seungjae Lee, Ruohan Gao, Furong Huang, Yiannis Aloimonos