Machine Learning
RewardUQ: A Unified Framework for Uncertainty-Aware Reward Models
Daniel Yang, Samuel Stante, Florian Redhardt et al. (+5 more)
Reward models are central to aligning large
language models (LLMs) with human preferences. Yet most approaches rely on pointwise reward estimates that overlook the epistemic uncertainty in reward models arising from limited human feedback. Recent work suggests that quantifying this uncertainty can reduce the costs of human annotation via uncertainty-guided active learning and mitigate reward overoptimization in LLM post-training. However, uncertainty-aware reward models have so far been adopted without thorough comparison, leaving them poorly understood. This work introduces a unified framework, RewardUQ, to systematically evaluate uncertainty quantification for reward models. We compare common methods along standard metrics measuring accuracy and calibration, and we propose a new ranking strategy incorporating both dimensions for a simplified comparison. Our experimental results suggest that model size and initialization have the most meaningful impact on performance, and most prior work could have benefited from alternative design choices. To foster the development and evaluation of new methods and aid the deployment in downstream applications, we release our open-source framework as a Python package. Our code is available at https://github.com/lasgroup/rewarduq.
Read the full abstract ▾
All authors Daniel Yang, Samuel Stante, Florian Redhardt, Lena Libon, Parnian Kassraie, Ido Hakimi, Barna Pásztor, Andreas Krause
Machine Learning
InfoNCE Induces Gaussian Distribution
Roy Betser, Eyal Gofer, Meir Yossef Levi et al. (+1 more)
Contrastive learning has become a cornerstone of modern representation learning, allowing training with massive unlabeled data for both task-specific and general (foundation) models. A prototypical loss in contrastive training is InfoNCE and its variants. In this work, we show that the InfoNCE objective induces Gaussian structure in representations that emerge from contrastive training. We establish this result in two complementary regimes. First, we show that under certain
alignment and concentration assumptions, projections of the high-dimensional representation asymptotically approach a multivariate Gaussian distribution. Next, under less strict assumptions, we show that adding a small asymptotically vanishing regularization term that promotes low feature norm and high feature
entropy leads to similar asymptotic results. We support our analysis with experiments on synthetic and CIFAR-10
datasets across multiple encoder architectures and sizes, demonstrating consistent Gaussian behavior. This perspective provides a principled explanation for commonly observed Gaussianity in contrastive representations. The resulting Gaussian model enables principled analytical treatment of learned representations and is expected to support a wide range of applications in contrastive learning.
Read the full abstract ▾
All authors Roy Betser, Eyal Gofer, Meir Yossef Levi, Guy Gilboa
Machine Learning · Cross-listed from cs.CR
Jailbreak Foundry: From Papers to Runnable Attacks for Reproducible Benchmarking
Zhicheng Fang, Jingjie Zheng, Chenxu Fu et al. (+1 more)
Jailbreak techniques for large
language models (LLMs) evolve faster than
benchmarks, making robustness estimates stale and difficult to compare across papers due to drift in
datasets, harnesses, and judging protocols. We introduce JAILBREAK FOUNDRY (JBF), a system that addresses this gap via a multi-
agent workflow to translate jailbreak papers into executable modules for immediate evaluation within a unified harness. JBF features three core components: (i) JBF-LIB for shared contracts and reusable utilities; (ii) JBF-FORGE for the multi-
agent paper-to-module translation; and (iii) JBF-EVAL for standardizing evaluations. Across 30 reproduced attacks, JBF achieves high fidelity with a mean (reproduced-reported) attack success rate (ASR) deviation of +0.26 percentage points. By leveraging shared infrastructure, JBF reduces attack-specific implementation code by more than half relative to original repositories and achieves an 82.5% mean reused-code ratio. This system enables a standardized AdvBench evaluation of all 30 attacks across 10 victim models using a consistent GPT-4o judge. By automating both attack integration and standardized evaluation, JBF offers a scalable solution for creating living
benchmarks that keep pace with the rapidly shifting security landscape.
Read the full abstract ▾
All authors Zhicheng Fang, Jingjie Zheng, Chenxu Fu, Wei Xu
Machine Learning · Cross-listed from q-bio.BM
Inference-time optimization for experiment-grounded protein ensemble generation
Advaith Maddipatla, Anar Rzayev, Marco Pegoraro et al. (+5 more)
Protein function relies on dynamic conformational ensembles, yet current generative models like AlphaFold3 often fail to produce ensembles that match experimental data. Recent experiment-guided generators attempt to address this by steering the reverse
diffusion process. However, these methods are limited by fixed sampling horizons and sensitivity to initialization, often yielding thermodynamically implausible results. We introduce a general inference-time optimization framework to solve these challenges. First, we optimize over latent representations to maximize ensemble log-likelihood, rather than perturbing structures post hoc. This approach eliminates dependence on
diffusion length, removes initialization bias, and easily incorporates external constraints. Second, we present novel sampling schemes for drawing Boltzmann-weighted ensembles. By combining structural priors from AlphaFold3 with force-field-based priors, we sample from their product distribution while balancing experimental likelihoods. Our results show that this framework consistently outperforms state-of-the-art guidance, improving diversity, physical energy, and agreement with data in X-ray crystallography and NMR, often fitting the experimental data better than deposited PDB structures. Finally, inference-time optimization experiments maximizing ipTM scores reveal that perturbing AlphaFold3
embeddings can artificially inflate model confidence. This exposes a vulnerability in current design metrics, whose mitigation could offer a pathway to reduce false discovery rates in binder engineering.
Read the full abstract ▾
All authors Advaith Maddipatla, Anar Rzayev, Marco Pegoraro, Martin Pacesa, Paul Schanda, Ailie Marx, Sanketh Vedula, Alex M. Bronstein
Machine Learning
Foundation World Models for Agents that Learn, Verify, and Adapt Reliably Beyond Static Environments
Florent Delgrange
The next generation of autonomous
agents must not only learn efficiently but also act reliably and adapt their behavior in open worlds. Standard approaches typically assume fixed tasks and environments with little or no novelty, which limits world models' ability to support
agents that must evolve their policies as conditions change. This paper outlines a
vision for foundation world models: persistent, compositional representations that unify reinforcement learning, reactive/program synthesis, and abstraction mechanisms. We propose an agenda built around four components: (i) learnable reward models from specifications to support optimization with clear objectives; (ii) adaptive formal verification integrated throughout learning; (iii) online abstraction calibration to quantify the reliability of the model's predictions; and (iv) test-time synthesis and world-model generation guided by verifiers. Together, these components enable
agents to synthesize verifiable programs, derive new policies from a small number of interactions, and maintain correctness while adapting to novelty. The resulting framework positions foundation world models as a substrate for learning,
reasoning, and adaptation, laying the groundwork for
agents that not only act well but can explain and justify the behavior they adopt.
Read the full abstract ▾
Machine Learning
MINT: Multimodal Imaging-to-Speech Knowledge Transfer for Early Alzheimer's Screening
Vrushank Ahire, Yogesh Kumar, Anouck Girard et al. (+1 more)
Alzheimer's disease is a progressive neurodegenerative disorder in which mild cognitive impairment (MCI) precedes dementia. Structural MRI provides biomarkers but requires costly infrastructure, limiting population-scale deployment.
Speech offers a non-invasive alternative, yet
speech-only classifiers are developed independently of neuroimaging and lack biological grounding for CN-versus-MCI classification. We propose MINT (
Multimodal Imaging-to-
Speech Knowledge Transfer), a three-stage framework that transfers MRI-derived biomarker structure to
speech during training. An MRI teacher defines a compact
embedding space for CN-versus-MCI classification, while a residual projection head aligns
speech representations to this space using a combined
geometric loss. The frozen MRI classifier enables imaging-free inference. On ADNI-4, aligned
speech achieves performance comparable to
speech baselines, while
multimodal fusion improves over MRI alone. Ablations identify dropout regularization and self-supervised pretraining as important design choices. To our knowledge, MINT is the first demonstration of MRI-to-
speech knowledge transfer for early Alzheimer's screening without imaging at inference.
Read the full abstract ▾
All authors Vrushank Ahire, Yogesh Kumar, Anouck Girard, M. A. Ganaie
Machine Learning
Intrinsic Lorentz Neural Network
Xianglong Shi, Ziheng Chen, Yunhan Jiang et al. (+1 more)
Real-world data frequently exhibit latent hierarchical structures, which can be naturally represented by
hyperbolic geometry. Although recent
hyperbolic neural networks have demonstrated promising results, many existing architectures remain partially intrinsic, mixing Euclidean operations with
hyperbolic ones or relying on extrinsic parameterizations. To address it, we propose the Intrinsic Lorentz Neural Network (ILNN), a fully intrinsic
hyperbolic architecture that conducts all computations within the Lorentz model. At its core, the network introduces a novel point-to-hyperplane fully connected layer (FC), replacing traditional Euclidean affine logits with closed-form
hyperbolic distances from features to learned Lorentz hyperplanes, thereby ensuring that the resulting
geometric decision functions respect the inherent
curvature. Around this fundamental layer, we design intrinsic modules: GyroLBN, a Lorentz batch normalization that couples gyro-centering with gyro-scaling, consistently outperforming both LBN and GyroBN while reducing training time. We additionally proposed a gyro-additive bias for the FC output, a Lorentz patch-concatenation operator that aligns the expected log-radius across feature blocks via a digamma-based scale, and a Lorentz dropout layer. Extensive experiments conducted on CIFAR-10/100 and two genomic
benchmarks (TEB and GUE) illustrate that ILNN achieves state-of-the-art performance and computational cost among
hyperbolic models and consistently surpasses strong Euclidean baselines. The code is available at \href{https://github.com/Longchentong/ILNN}{\textcolor{magenta}{this url}}.
Read the full abstract ▾
All authors Xianglong Shi, Ziheng Chen, Yunhan Jiang, Nicu Sebe
Machine Learning
Learning Generation Orders for Masked Discrete Diffusion Models via Variational Inference
David Fox, Sam Bowyer, Song Liu et al. (+3 more)
Masked discrete
diffusion models (MDMs) are a promising new approach to generative modelling, offering the ability for parallel
token generation and therefore greater efficiency than autoregressive counterparts. However, achieving an optimal balance between parallel generation and sample quality remains an open problem. Current approaches primarily address this issue through fixed, heuristic parallel sampling methods. There exist some recent learning based approaches to this problem, but its formulation from the perspective of variational inference remains underexplored. In this work, we propose a variational inference framework for learning parallel generation orders for MDMs. As part of our method, we propose a parameterisation for the approximate posterior of generation orders which facilitates parallelism and efficient sampling during training. Using this method, we conduct preliminary experiments on the GSM8K
dataset, where our method performs competitively against heuristic sampling strategies in the regime of highly parallel generation. For example, our method achieves 33.1% accuracy with an average of only only 4 generation steps, compared to 23.7-29.0% accuracy achieved by standard competitor methods in the same number of steps. We believe further experiments and analysis of the method will yield valuable insights into the problem of parallel generation with MDMs.
Read the full abstract ▾
All authors David Fox, Sam Bowyer, Song Liu, Laurence Aitchison, Raul Santos-Rodriguez, Mengyue Yang
Machine Learning
Hierarchical Concept-based Interpretable Models
Oscar Hill, Mateo Espinosa Zarlenga and Mateja Jamnik
Modern deep neural networks remain challenging to interpret due to the opacity of their latent representations, impeding model understanding, debugging, and debiasing. Concept
Embedding Models (CEMs) address this by mapping inputs to human-interpretable concept representations from which tasks can be predicted. Yet, CEMs fail to represent inter-concept relationships and require concept annotations at different granularities during training, limiting their applicability. In this paper, we introduce Hierarchical Concept
Embedding Models (HiCEMs), a new family of CEMs that explicitly model concept relationships through hierarchical structures. To enable HiCEMs in real-world settings, we propose Concept Splitting, a method for automatically discovering finer-grained sub-concepts from a pretrained CEM's
embedding space without requiring additional annotations. This allows HiCEMs to generate fine-grained explanations from limited concept labels, reducing annotation burdens. Our evaluation across multiple
datasets, including a user study and experiments on PseudoKitchens, a newly proposed concept-based
dataset of 3D kitchen renders, demonstrates that (1) Concept Splitting discovers human-interpretable sub-concepts absent during training that can be used to train highly accurate HiCEMs, and (2) HiCEMs enable powerful test-time concept interventions at different granularities, leading to improved task accuracy.
Read the full abstract ▾
Machine Learning · Cross-listed from cs.CL
Benchmarking BERT-based Models for Sentence-level Topic Classification in Nepali Language
Nischal Karki, Bipesh Subedi, Prakash Poudyal et al. (+2 more)
Transformer-based models such as BERT have significantly advanced Natural
Language Processing (NLP) across many
languages. However, Nepali, a low-resource
language written in Devanagari script, remains relatively underexplored. This study
benchmarks multilingual, Indic, Hindi, and Nepali BERT variants to evaluate their effectiveness in Nepali topic classification. Ten pre-trained models, including mBERT, XLM-R, MuRIL, DevBERT, HindiBERT, IndicBERT, and NepBERTa, were fine-tuned and tested on the balanced Nepali
dataset containing 25,006 sentences across five conceptual domains and the performance was evaluated using accuracy, weighted precision, recall, F1-score, and AUROC metrics. The results reveal that Indic models, particularly MuRIL-large, achieved the highest F1-score of 90.60%, outperforming
multilingual and monolingual models. NepBERTa also performed competitively with an F1-score of 88.26%. Overall, these findings establish a robust baseline for future document-level classification and broader Nepali NLP applications.
Read the full abstract ▾
All authors Nischal Karki, Bipesh Subedi, Prakash Poudyal, Rupak Raj Ghimire, Bal Krishna Bal
Machine Learning · Cross-listed from cs.RO
Learning to Build: Autonomous Robotic Assembly of Stable Structures Without Predefined Plans
Jingwen Wang, Johannes Kirschner, Paul Rolland et al. (+2 more)
This paper presents a novel autonomous robotic assembly framework for constructing stable structures without relying on predefined architectural blueprints. Instead of following fixed plans, construction tasks are defined through targets and obstacles, allowing the system to adapt more flexibly to environmental uncertainty and variations during the building process. A reinforcement learning (RL) policy, trained using deep Q-learning with successor features, serves as the decision-making component. As a proof of concept, we evaluate the approach on a
benchmark of 15 2D robotic assembly tasks of discrete block construction. Experiments using a real-world closed-loop robotic setup demonstrate the feasibility of the method and its ability to handle construction noise. The results suggest that our framework offers a promising direction for more adaptable and robust robotic construction in real-world environments.
Read the full abstract ▾
All authors Jingwen Wang, Johannes Kirschner, Paul Rolland, Luis Salamanca, Stefana Parascho
Machine Learning · Cross-listed from cs.CV
Half-Truths Break Similarity-Based Retrieval
Bora Kargi, Arnas Uselis and Seong Joon Oh
When a text description is extended with an additional detail, image-text similarity should drop if that detail is wrong. We show that CLIP-style dual encoders often violate this intuition: appending a plausible but incorrect object or relation to an otherwise correct description can increase the similarity score. We call such cases half-truths. On COCO, CLIP prefers the correct shorter description only 34.0% of the time, and performance drops to 27.8% when the added detail is a relation. We trace this vulnerability to weak supervision on caption parts: contrastive training aligns full sentences but does not explicitly enforce that individual entities and relations are grounded. We propose CS-CLIP (Component-Supervised CLIP), which decomposes captions into entity and relation units, constructs a minimally edited foil for each unit, and fine-tunes the model to score the correct unit above its foil while preserving standard dual-encoder inference. CS-CLIP raises half-truth accuracy to 76.4% and improves average performance on established compositional
benchmarks by 5.8 points, suggesting that reducing half-truth errors aligns with broader gains in compositional understanding.
Read the full abstract ▾
Machine Learning · Cross-listed from cs.CV
SegMate: Asymmetric Attention-Based Lightweight Architecture for Efficient Multi-Organ Segmentation
Andrei-Alexandru Bunea, Dan-Matei Popovici and Radu Tudor Ionescu
State-of-the-art models for medical image
segmentation achieve excellent accuracy, but require substantial computational resources, limiting deployment in typical resource-constrained clinical settings. We present SegMate, an efficient 2.5D framework that achieves state-of-the-art accuracy, while considerably reducing computational requirements. Our efficient design is the result of meticulously integrating asymmetric architectures,
attention mechanisms, multi-scale feature fusion, slice-based positional conditioning, and multi-task optimization. We demonstrate the efficiency-accuracy trade-off of our framework across three modern backbones (EfficientNetV2-M, MambaOut-Tiny, FastViT-T12). We perform experiments on three
datasets: TotalSegmentator, SegTHOR and AMOS22. Compared with the vanilla models, SegMate reduces computation (GFLOPs) by up to 2.5x and
memory footprint (VRAM) by up to 2.1x, while generally registering performance gains of around 1%. On TotalSegmentator, we achieve a Dice score of 93.51% with only 295 MB peak GPU
memory. Zero-shot cross-
dataset evaluations on SegTHOR and AMOS22 demonstrate strong generalization, with Dice scores of up to 86.85% and 89.35%, respectively. SegMate variants in this study empower single GPU setups to process up to 10,000 volumes per hour with just 74 J per volume using batch inference, enabling population-scale CT processing on commodity hardware. We release our open-source code at https://github.com/andreibunea99/SegMate.
Read the full abstract ▾
Machine Learning · Cross-listed from cs.CV
Experience-Guided Self-Adaptive Cascaded Agents for Breast Cancer Screening and Diagnosis with Reduced Biopsy Referrals
Pramit Saha, Mohammad Alsharid, Joshua Strong et al. (+1 more)
We propose an experience-guided cascaded multi-
agent framework for Breast Ultrasound Screening and Diagnosis, called BUSD-
Agent, that aims to reduce diagnostic escalation and unnecessary biopsy referrals. Our framework models screening and diagnosis as a two-stage, selective decision-making process. A lightweight `screening clinic'
agent, restricted to classification models as tools, selectively filters out benign and normal cases from further diagnostic escalation when malignancy risk and uncertainty are estimated as low. Cases that have higher risks are escalated to the `diagnostic clinic'
agent, which integrates richer perception and radiological description tools to make a secondary decision on biopsy referral. To improve
agent performance, past records of pathology-confirmed outcomes along with image
embeddings, model predictions, and historical
agent actions are stored in a
memory bank as structured decision trajectories. For each new case, BUSD-
Agent retrieves similar past cases based on image, model response and confidence similarity to condition the
agent's current decision policy. This enables
retrieval-conditioned in-context adaptation that dynamically adjusts model trust and escalation thresholds from prior experiences without parameter updates. Evaluation across 10 breast ultrasound
datasets shows that the proposed experience-guided workflow reduces diagnostic escalation in BUSD-
Agent from 84.95% to 58.72% and overall biopsy referrals from 59.50% to 37.08%, compared to the same architecture without trajectory conditioning, while improving average screening specificity by 68.48% and diagnostic specificity by 6.33%.
Read the full abstract ▾
All authors Pramit Saha, Mohammad Alsharid, Joshua Strong, J. Alison Noble
Machine Learning
LK Losses: Direct Acceptance Rate Optimization for Speculative Decoding
Alexander Samarin, Sergei Krutikov, Anton Shevtsov et al. (+3 more)
Speculative decoding accelerates autoregressive large
language model (LLM) inference by using a lightweight draft model to propose candidate
tokens that are then verified in parallel by the target model. The speedup is significantly determined by the acceptance rate, yet standard training minimizes Kullback-Leibler (KL) divergence as a proxy objective. While KL divergence and acceptance rate share the same global optimum, small draft models, having limited capacity, typically converge to suboptimal solutions where minimizing KL does not guarantee maximizing acceptance rate. To address this issue, we propose LK losses, special training objectives that directly target acceptance rate. Comprehensive experiments across four draft architectures and six target models, ranging from 8B to 685B parameters, demonstrate consistent improvements in acceptance metrics across all configurations compared to the standard KL-based training. We evaluate our approach on general, coding and math domains and report gains of up to 8-10% in average acceptance length. LK losses are easy to implement, introduce no computational overhead and can be directly integrated into any existing speculator training framework, making them a compelling alternative to the existing draft training objectives.
Read the full abstract ▾
All authors Alexander Samarin, Sergei Krutikov, Anton Shevtsov, Sergei Skvortsov, Filipp Fisin, Alexander Golubev
Machine Learning
A Theory of Random Graph Shift in Truncated-Spectrum vRKHS
Zhang Wan, Tingting Mu and Samuel Kaski
This paper develops a theory of graph classification under domain shift through a random-graph generative lens, where we consider intra-class graphs sharing the same random graph model (RGM) and the domain shift induced by changes in RGM components. While classic domain adaptation (DA) theories have well-underpinned existing techniques to handle graph distribution shift, the information of graph samples, which are itself structured objects, is less explored. The non-Euclidean nature of graphs and specialized architectures for graph learning further complicate a fine-grained analysis of graph distribution shifts. In this paper, we propose a theory that assumes RGM as the data generative process, exploiting its connection to hypothesis complexity in function space perspective for such fine-grained analysis. Building on a vector-valued reproducing kernel
Hilbert space (vRKHS) formulation, we derive a generalization bound whose shift penalty admits a factorization into (i) a domain discrepancy term, (ii) a spectral-geometry term summarized by the accessible truncated spectrum, and (iii) an amplitude term that aggregates convergence and construction-stability effects. We empirically verify the insights on these terms in both real data and simulations.
Read the full abstract ▾
Machine Learning · Cross-listed from cs.AI
RF-Agent: Automated Reward Function Design via Language Agent Tree Search
Ning Gao, Xiuhui Zhang, Xingyu Jiang et al. (+3 more)
Designing efficient reward functions for low-level control tasks is a challenging problem. Recent research aims to reduce reliance on expert experience by using Large
Language Models (LLMs) with task information to generate dense reward functions. These methods typically rely on training results as feedback, iteratively generating new reward functions with greedy or evolutionary
algorithms. However, they suffer from poor utilization of historical feedback and inefficient search, resulting in limited improvements in complex control tasks. To address this challenge, we propose RF-
Agent, a framework that treats LLMs as
language agents and frames reward function design as a sequential decision-making process, enhancing optimization through better contextual
reasoning. RF-
Agent integrates Monte Carlo Tree Search (MCTS) to manage the reward design and optimization process, leveraging the multi-stage contextual
reasoning ability of LLMs. This approach better utilizes historical information and improves search efficiency to identify promising reward functions. Outstanding experimental results in 17 diverse low-level control tasks demonstrate the effectiveness of our method. The source code is available at https://github.com/deng-ai-lab/RF-
Agent.
Read the full abstract ▾
All authors Ning Gao, Xiuhui Zhang, Xingyu Jiang, Mukang You, Mohan Zhang, Yue Deng
Machine Learning · Cross-listed from cs.CR
Exploring Robust Intrusion Detection: A Benchmark Study of Feature Transferability in IoT Botnet Attack Detection
Alejandro Guerra-Manzanares and Jialin Huang
Cross-domain intrusion detection remains a critical challenge due to significant variability in network traffic characteristics and feature distributions across environments. This study evaluates the transferability of three widely used flow-based feature sets (Argus, Zeek and CICFlowMeter) across four widely used
datasets representing heterogeneous IoT and Industrial IoT network conditions. Through extensive experiments, we evaluate in- and cross-domain performance across multiple classification models and analyze feature importance using SHapley Additive exPlanations (SHAP). Our results show that models trained on one domain suffer significant performance degradation when applied to a different target domain, reflecting the sensitivity of IoT intrusion detection systems to distribution shifts. Furthermore, the results evidence that the choice of classification
algorithm and feature representations significantly impact transferability. Beyond reporting performance differences and thorough analysis of the transferability of features and feature spaces, we provide practical guidelines for feature engineering to improve robustness under domain variability. Our findings suggest that effective intrusion detection requires both high in-domain performance and resilience to cross-domain variability, achievable through careful feature space design, appropriate
algorithm selection and adaptive strategies.
Read the full abstract ▾
Machine Learning · Cross-listed from cs.CV
Bandwidth-adaptive Cloud-Assisted 360-Degree 3D Perception for Autonomous Vehicles
Faisal Hawladera, Rui Meireles, Gamal Elghazaly et al. (+2 more)
A key challenge for autonomous driving lies in maintaining real-time situational awareness regarding surrounding obstacles under strict latency constraints. The high processing requirements coupled with limited onboard computational resources can cause delay issues, particularly in complex urban settings. To address this, we propose leveraging Vehicle-to-Everything (V2X) communication to partially offload processing to the cloud, where compute resources are abundant, thus reducing overall latency. Our approach utilizes
transformer-based models to fuse multi-camera sensor data into a comprehensive Bird's-Eye View (BEV) representation, enabling accurate 360-degree 3D object detection. The computation is dynamically split between the vehicle and the cloud based on the number of layers processed locally and the quantization level of the features. To further reduce network load, we apply feature vector clipping and compression prior to transmission. In a real-world experimental evaluation, our hybrid strategy achieved a 72 % reduction in end-to-end latency compared to a traditional onboard solution. To adapt to fluctuating network conditions, we introduce a dynamic optimization
algorithm that selects the split point and quantization level to maximize detection accuracy while satisfying real-time latency constraints. Trace-based evaluation under realistic bandwidth variability shows that this adaptive approach improves accuracy by up to 20 % over static parameterization with the same latency performance.
Read the full abstract ▾
All authors Faisal Hawladera, Rui Meireles, Gamal Elghazaly, Ana Aguiar, Raphaël Frank
Machine Learning · Cross-listed from math.OC
A distributed semismooth Newton based augmented Lagrangian method for distributed optimization
Qihao Ma, Chengjing Wang, Peipei Tang et al. (+2 more)
This paper proposes a novel distributed semismooth Newton based augmented Lagrangian method for solving a class of optimization problems over networks, where the global objective is defined as the sum of locally held cost functions, and communication is restricted to neighboring
agents. Specifically, we employ the augmented Lagrangian method to solve an equivalently reformulated constrained version of the original problem. Each resulting subproblem is solved inexactly via a distributed semismooth Newton method. By fully leveraging the structure of the generalized Hessian, a distributed accelerated proximal gradient method is proposed to compute the Newton direction efficiently, eliminating the need to communicate with full Hessian
matrices. Theoretical results are also obtained to guarantee the convergence of the proposed
algorithm. Numerical experiments demonstrate the efficiency and superiority of our
algorithm compared to state-of-the-art distributed
algorithms.
Read the full abstract ▾
All authors Qihao Ma, Chengjing Wang, Peipei Tang, Dunbiao Niu, Aimin Xu
Machine Learning
ULW-SleepNet: An Ultra-Lightweight Network for Multimodal Sleep Stage Scoring
Zhaowen Wang, Dongdong Zhou, Qi Xu et al. (+3 more)
Automatic sleep stage scoring is crucial for the diagnosis and treatment of sleep disorders. Although deep learning models have advanced the field, many existing models are computationally demanding and designed for single-channel electroencephalography (
EEG), limiting their practicality for
multimodal polysomnography (PSG) data. To overcome this, we propose ULW-SleepNet, an ultra-lightweight
multimodal sleep stage scoring framework that efficiently integrates information from multiple physiological signals. ULW-SleepNet incorporates a novel Dual-Stream Separable Convolution (DSSC) Block, depthwise separable convolutions, channel-wise parameter sharing, and global average pooling to reduce computational overhead while maintaining competitive accuracy. Evaluated on the Sleep-EDF-20 and Sleep-EDF-78
datasets, ULW-SleepNet achieves accuracies of 86.9% and 81.4%, respectively, with only 13.3K parameters and 7.89M FLOPs. Compared to state-of-the-art methods, our model reduces parameters by up to 98.6% with only marginal performance loss, demonstrating its strong potential for real-time sleep monitoring on wearable and IoT devices. The source code for this study is publicly available at https://github.com/wzw999/ULW-SLEEPNET.
Read the full abstract ▾
All authors Zhaowen Wang, Dongdong Zhou, Qi Xu, Fengyu Cong, Mohammad Al-Sa'd, Jenni Raitoharju
Machine Learning · Cross-listed from cs.CR
MI$^2$DAS: A Multi-Layer Intrusion Detection Framework with Incremental Learning for Securing Industrial IoT Networks
Wei Lian and Alejandro Guerra-Manzanares
The rapid expansion of Industrial IoT (IIoT) systems has amplified security challenges, as heterogeneous devices and dynamic traffic patterns increase exposure to sophisticated and previously unseen cyberattacks. Traditional intrusion detection systems often struggle in such environments due to their reliance on extensive labeled data and limited ability to detect new threats. To address these challenges, we propose MI$^2$DAS, a multi-layer intrusion detection framework that integrates anomaly-based hierarchical traffic pooling, open-set recognition to distinguish between known and unknown attacks and incremental learning for adapting to novel attack types with minimal labeling. Experiments conducted on the Edge-IIoTset
dataset demonstrate strong performance across all layers. In the first layer, GMM achieves superior normal-attack discrimination (accuracy = 0.953, TPR = 1.000). In open-set recognition, GMM attains a recall of 0.813 for known attacks, while LOF achieves 0.882 recall for unknown attacks. For fine-grained classification of known attacks, Random Forest achieves a macro-F1 of 0.941. Finally, the incremental learning module maintains robust performance when incorporation novel attack classes, achieving a macro-F1 of 0.8995. These results showcase MI$^2$DAS as an effective, scalable and adaptive framework for enhancing IIoT security against evolving threats.
Read the full abstract ▾
Machine Learning · Cross-listed from cs.CR
Enhancing Continual Learning for Software Vulnerability Prediction: Addressing Catastrophic Forgetting via Hybrid-Confidence-Aware Selective Replay for Temporal LLM Fine-Tuning
Xuhui Dou, Hayretdin Bahsi and Alejandro Guerra-Manzanares
Recent work applies Large
Language Models (LLMs) to source-code vulnerability detection, but most evaluations still rely on random train-test splits that ignore time and overestimate real-world performance. In practice, detectors are deployed on evolving code bases and must recognise future vulnerabilities under temporal distribution shift. This paper investigates continual fine-tuning of a decoder-style
language model (microsoft/phi-2 with LoRA) on a CVE-linked
dataset spanning 2018-2024, organised into bi-monthly windows. We evaluate eight continual learning strategies, including window-only and cumulative training, replay-based baselines and regularisation-based variants. We propose Hybrid Class-Aware Selective Replay (Hybrid-CASR), a confidence-aware replay method for binary vulnerability classification that prioritises uncertain samples while maintaining a balanced ratio of VULNERABLE and FIXED functions in the replay buffer. On bi-monthly forward evaluation Hybrid-CASR achieves a Macro-F1 of 0.667, improving on the window-only baseline (0.651) by 0.016 with statistically significant gains ($p = 0.026$) and stronger backward retention (IBR@1 of 0.741). Hybrid-CASR also reduces training time per window by about 17 percent compared to the baseline, whereas cumulative training delivers only a minor F1 increase (0.661) at a 15.9-fold computational cost. Overall, the results show that selective replay with class balancing offers a practical accuracy-efficiency trade-off for LLM-based temporal vulnerability detection under continuous temporal drift.
Read the full abstract ▾
Machine Learning
FedNSAM:Consistency of Local and Global Flatness for Federated Learning
Junkang Liu, Fanhua Shang, Yuxuan Tian et al. (+2 more)
In
federated learning (FL), multi-step local updates and data heterogeneity usually lead to sharper global minima, which degrades the performance of the global model. Popular FL
algorithms integrate sharpness-aware minimization (SAM) into local training to address this issue. However, in the high data heterogeneity setting, the flatness in local training does not imply the flatness of the global model. Therefore, minimizing the sharpness of the local loss surfaces on the client data does not enable the effectiveness of SAM in FL to improve the generalization ability of the global model. We define the flatness distance to explain this phenomenon. By rethinking the SAM in FL and theoretically analyzing the flatness distance, we propose a novel FedNSAM
algorithm that accelerates the SAM
algorithm by introducing global Nesterov momentum into the local update to harmonize the consistency of global and local flatness. FedNSAM uses the global Nesterov momentum as the direction of local estimation of client global perturbations and extrapolation. Theoretically, we prove a tighter convergence bound than FedSAM by Nesterov extrapolation. Empirically, we conduct comprehensive experiments on CNN and
Transformer models to verify the superior performance and efficiency of FedNSAM. The code is available at https://github.com/junkangLiu0/FedNSAM.
Read the full abstract ▾
All authors Junkang Liu, Fanhua Shang, Yuxuan Tian, Hongying Liu, Yuanyuan Liu
Machine Learning · Cross-listed from cs.CL
Sebastian Gerstner, Leonor Veloso and Hinrich Schütze
We present **GLUScope**, an open-source resource for analyzing
neurons in
Transformer-based
language models (LMs), intended for interpretability researchers. We focus on the now-standard *GLU-based activation functions* such as SwiGLU, used in virtually all recent open-weights LMs. This introduces a new challenge: understanding positive activations is not enough. Instead, both the *gate* and the *in* activation of a
neuron can be positive or negative, leading to four different possible sign combinations (*activation quadrants*) that in some cases have quite different functionalities (e.g. promoting vs. suppressing a specific set of
tokens). Accordingly, for any
neuron, our tool shows text examples for each of the four sign combinations, and indicates how often each combination occurs. Our resource can lead to novel insights: in particular, thanks to analyzing all four activation quadrants, it can surface interpretable behavior that other tools would miss. We also release the code used to create the resource, thus enabling researchers to build on it by adding more models, text corpora, or
neuron pages.
Read the full abstract ▾
Machine Learning
Inferring Chronic Treatment Onset from ePrescription Data: A Renewal Process Approach
Pavlin G. Poličar, Dalibor Stanimirović and Blaž Zupan
Longitudinal electronic health record (EHR) data are often left-censored, making diagnosis records incomplete and unreliable for determining disease onset. In contrast, outpatient prescriptions form renewal-based trajectories that provide a continuous signal of disease management. We propose a probabilistic framework to infer chronic treatment onset by modeling prescription dynamics as a renewal process and detecting transitions from sporadic to sustained therapy via change-point detection between a baseline Poisson (sporadic prescribing) regime and a regime-specific Weibull (sustained therapy) renewal model. Using a nationwide ePrescription
dataset of 2.4 million individuals, we show that the approach yields more temporally plausible onset estimates than naive rule-based triggering, substantially reducing implausible early detections under strong left censoring. Detection performance varies across diseases and is strongly associated with prescription density, highlighting both the strengths and limits of treatment-based onset inference.
Read the full abstract ▾
Machine Learning
Learning to maintain safety through expert demonstrations in settings with unknown constraints: A Q-learning perspective
George Papadopoulos and George A. Vouros
Given a set of trajectories demonstrating the execution of a task safely in a constrained MDP with observable rewards but with unknown constraints and non-observable costs, we aim to find a policy that maximizes the likelihood of demonstrated trajectories trading the balance between being conservative and increasing significantly the likelihood of high-rewarding trajectories but with potentially unsafe steps. Having these objectives, we aim towards learning a policy that maximizes the probability of the most $promising$ trajectories with respect to the demonstrations. In so doing, we formulate the “promise" of individual state-action pairs in terms of $Q$ values, which depend on task-specific rewards as well as on the assessment of states' safety, mixing expectations in terms of rewards and safety. This entails a safe Q-learning perspective of the inverse learning problem under constraints: The devised Safe $Q$ Inverse Constrained Reinforcement Learning (SafeQIL)
algorithm is compared to state-of-the art inverse constraint reinforcement learning
algorithms to a set of challenging
benchmark tasks, showing its merits.
Read the full abstract ▾
Machine Learning
Beyond State-Wise Mirror Descent: Offline Policy Optimization with Parametric Policies
Xiang Li, Yuheng Zhang and Nan Jiang
We investigate the theoretical aspects of offline reinforcement learning (RL) under general function approximation. While prior works (e.g., Xie et al., 2021) have established the theoretical foundations of learning a good policy from offline data via pessimism, existing
algorithms that are computationally tractable (often in an oracle-efficient sense), such as PSPI, only apply to finite and small action spaces. Moreover, these
algorithms rely on state-wise mirror descent and require actors to be implicitly induced from the critic functions, failing to accommodate standalone policy parameterization which is ubiquitous in practice. In this work, we address these limitations and extend the theoretical guarantees to parameterized policy classes over large or continuous action spaces. When extending mirror descent to parameterized policies, we identify contextual coupling as the core difficulty, and show how connecting mirror descent to natural policy gradient leads to novel analyses, guarantees, and algorithmic insights, including a surprising unification between offline RL and imitation learning.
Read the full abstract ▾
Machine Learning · Cross-listed from cs.CY
ReasonX: Declarative Reasoning on Explanations
Laura State, Salvatore Ruggieri and Franco Turini
Explaining opaque Machine Learning (ML) models has become an increasingly important challenge. However, current eXplanation in AI (XAI) methods suffer several shortcomings, including insufficient abstraction, limited user interactivity, and inadequate integration of symbolic knowledge. We propose ReasonX, an explanation tool based on expressions (or, queries) in a closed
algebra of operators over theories of linear constraints. ReasonX provides declarative and interactive explanations for decision trees, which may represent the ML models under analysis or serve as global or local surrogate models for tabular-data black-box classifiers. Users can express background or common sense knowledge as linear constraints. This allows for
reasoning at multiple levels of abstraction, ranging from fully specified examples to under-specified or partially constrained ones. ReasonX leverages Mixed-Integer Linear Programming (MILP) to reason over the features of factual and contrastive instances. We present here the architecture of ReasonX, which consists of a Python layer, closer to the user, and a Constraint Logic Programming (CLP) layer, which implements a meta-interpreter of the query
algebra. The capabilities of ReasonX are demonstrated through qualitative examples, and compared to other XAI tools through quantitative experiments.
Read the full abstract ▾
Machine Learning
Actor-Critic Pretraining for Proximal Policy Optimization
Andreas Kernbach, Amr Elsheikh, Nicolas Grupp et al. (+2 more)
Reinforcement learning (RL) actor-critic
algorithms enable autonomous learning but often require a large number of environment interactions, which limits their applicability in robotics. Leveraging expert data can reduce the number of required environment interactions. A common approach is actor pretraining, where the actor network is initialized via behavioral cloning on expert demonstrations and subsequently fine-tuned with RL. In contrast, the initialization of the critic network has received little
attention, despite its central role in policy optimization. This paper proposes a pretraining approach for actor-critic
algorithms like Proximal Policy Optimization (PPO) that uses expert demonstrations to initialize both networks. The actor is pretrained via behavioral cloning, while the critic is pretrained using returns obtained from rollouts of the pretrained policy. The approach is evaluated on 15 simulated robotic manipulation and locomotion tasks. Experimental results show that actor-critic pretraining improves sample efficiency by 86.1% on average compared to no pretraining and by 30.9% to actor-only pretraining.
Read the full abstract ▾
All authors Andreas Kernbach, Amr Elsheikh, Nicolas Grupp, René Nagel, Marco F. Huber