cs.LG · 2026-08-13 · No. 83
Machine Learning, 2026-08-13.
65 new papers in cs.LG. Titles, authors,
abstracts. Links to arXiv. Want this in your inbox every morning? Subscribe →
01 — The papers
65 entries-
01
AI4AI at Test-Time: Strong-to-Weak Capability Transfer via Harnesses
Cheng Qian, Wenting Zhao, Liangwei Yang, Heng Wang, Jielin Qiu, Heng Ji, Silvio Savarese, Huan Wang, Shelby Heinecke
cs.LG · cs.AI · cs.CL
Recent work on distillation transfers the capabilities of large models to smaller ones often by updating the latter's parameters, through teacher forcing, on-policy distillation, and related training-time methods. In this paper, we ask whether such transfer can instead occur at test time. We study strong-to-weak scaffolding: whether a stronger builder model can construct inference-time harnesses that help a weaker target model solve tasks...
-
02
Redistribution-based Cost Inference Improves Sparse Safe Offline RL
Ebenezer Gelo, Geraud Nangue Tasse, Steven James, Benjamin Rosman
cs.LG · cs.AI
Safe offline RL typically assumes access to dense per-step cost annotations, but in practice supervisors provide only trajectory-level stop-feedback: a binary signal at the first unsafe transition, with no per-step attribution. We frame this as a temporal credit assignment problem and propose the Redistribution-based Cost Inference (RCI) framework, which converts sparse stop-feedback into dense per-step costs via return decomposition, then...
-
03
A Framework for Designing Reward Functions: From Objectives to Features to Human-Aligned Reward Functions
Di Yang Shi, W. Bradley Knox
cs.LG
We present a formal process to enable non-experts to instantiate and iterate on human-aligned reward functions, i.e. reward functions that adhere to a given preference ordering over trajectories. Given a task described in natural language, our process produces a linear reward function in three steps: distill the task's objectives into a set of fundamental objectives and derive measurable outcome variables that capture those fundamental...
-
04
Earth observation embeddings are effective sub-grid descriptors for probabilistic weather downscaling
Pedro Sousa, Will Tebbutt, Sadiq Jaffer, Robin Young, Anil Madhavapeddy, Richard E. Turner
cs.LG · physics.ao-ph
Global weather reanalyses and forecasts resolve the evolving atmospheric state on coarse grids, but site-specific applications require predictions at arbitrary locations where near-surface conditions also depend on unresolved terrain and land-surface properties. Existing probabilistic downscalers address this gap using hand-crafted topographic descriptors. We ask instead whether Earth observation foundation models can provide transferable...
-
05
Calibration Bets on the Past: Post-Training Quantization for Financial Time-Series Forecasting
Junyi Ye, Ivy Gateri Wanjiku
cs.LG · q-fin.ST
Financial forecasting models are typically developed in full precision, yet production deployment often requires low-precision inference to reduce memory and computational cost. Post-training quantization (PTQ) enables such deployment without retraining. However, reliable activation quantization requires calibration: activation ranges are estimated from historical data before deployment and then remain fixed during future inference. The...
-
06
An Efficient Near-Optimal Algorithm for Adversarial $m$-Set Bandits
Francesco Bacchiocchi, Tommaso Cesari, Roberto Colomboni
cs.LG
We study adversarial combinatorial bandits with $m$-set actions, where at each round the learner selects $m$ out of $d$ items and observes only the aggregate loss of the selected items. The resulting action set contains $K=\binom{d}{m}$ elements and can therefore be exponentially large. Nevertheless, the loss of every action is determined by the same $d$-dimensional vector of item losses. We propose a computationally efficient algorithm that...
-
07
ScreenShot: A Foundation Model for Few-Shot Combination Drug Screening
Antoine de Mathelin, Christopher Tosh, Wesley Tansey
cs.LG
Treating patients with combinations of drugs reduces the risk of resistance to any individual drug. Finding effective combinations is difficult because the large search space makes combinatorial screens prohibitively expensive, time consuming, and often technically infeasible. Predictive models can fill this gap, yet existing methods typically require molecular profiling of each sample and per-cohort training, limiting their applicability...
-
08
HYDRA: Hyperbolic Dynamic Representation Architecture for Kolmogorov-Arnold Networks
Zhao Su, Yuxin Xia, Haoran Li, Jun Shen, Qi Zhu, Qingguo Zhou, Binbin Yong
cs.LG · cs.AI
Kolmogorov-Arnold Networks (KANs) enhance nonlinear function approximation by replacing scalar weights with learnable univariate functions. However, assigning an independent function to every connection results in substantial parameter redundancy, limiting their scalability and efficiency. To reduce this redundancy, we introduce \textbf{HY}perbolic \textbf{D}ynamic \textbf{R}epresentation \textbf{A}rchitecture (HYDRA), a parameter-efficient...
-
09
Adversarial Resilience of Poisson-Process Submodular Maximization over Matroids: From Robust Offline Optimization to Full-Bandit Learning
Vaneet Aggarwal
cs.LG · cs.AI · cs.CC · math.OC
We study nonnegative submodular maximization subject to a general matroid when the offline algorithm is given an arbitrary controlled value oracle. Our main result is an adversarial resilience theorem for the Spiteful Greedy Swap Poisson Process (SGS-Poisson): without modifying its Poisson intensity, single-element exchange rule, or spiteful drop step, the algorithm retains limiting approximation factors $1/e$ for non-monotone objectives and...
-
10
Attractor Image-Based Deep Learning of Arterial Pulse Waves for Age Classification
Sara Vardanega, Patrick Segers, Philip Aston, Ernst Rietzschel, Jordi Alastruey, Manasi Nandi
cs.LG
Arterial pulse waveform morphology evolves with age, reflecting structural and functional changes in the cardiovascular system. Thus, vascular age is a valuable surrogate marker of cardiovascular health, and premature vascular ageing can indicate increased disease risk. Pulse wave analysis could support risk stratification in otherwise asymptomatic adults. We transformed pulse wave time-series data from photoplethysmography (PPG) and arterial...
-
11
Beyond Parameter Space: NTK-Guided Personalized Aggregation for Robust Federated Learning
Mirko Konstantin, Stefan Zachow, Anirban Mukhopadhyay
cs.LG
Federated learning (FL) enables collaborative model training across distributed clients while keeping data local. A central challenge is determining which client updates are beneficial for aggregation with respect to each client's target domain. Existing methods typically address this problem in parameter space by comparing model parameters or gradients. However, parameter-space similarity can be a poor proxy for predictive behavior,...
-
12
Confidence Calibration of Deep Learning Systems
Coby Penso
cs.LG · cs.AI · stat.ML
In high-stakes applications, reliable confidence estimates are as important as the predictions themselves. Confidence calibration ensures that predicted probabilities reflect the likelihood of correctness, making it essential for safe deployment of deep learning models. However, existing methods typically assume access to clean validation data, which is often unrealistic due to label noise and domain shifts. This thesis develops methods for...
-
13
Task- and dataset-specific information in protein language models
Roman Joeres, Ilya Senatorov, Olga V. Kalinina
cs.LG · q-bio.BM
Protein language models (PLMs) have transferred the latest advances from natural language processing to computational biology. These models, trained on large corpora of protein sequence data, are widely used to translate amino acid sequences into latent-space embeddings, ready for use in diverse downstream tasks (DTs). By a common consensus, embeddings from the model's last layer are used, and the model's internal behavior remains poorly...
-
14
NAE: Normalizing AutoEncoder
Muhammad Abdur Rafae, Niels Landwehr
cs.LG
We consider the setting of Normalizing flows with approximate inverses, an established paradigm spanning both full-dimensional ($d=D$) and bottleneck ($d<D$) settings, and group these models under the term flow autoencoders. We present a theoretical investigation into their training dynamics and prove that the proposed loss used by existing approaches is suboptimal; specifically, both encoder and decoder surrogates must be optimized in...
-
15
Faithful, Sufficient and Understandable: Rethinking Graph Counterfactual Explanations via Discrete Diffusion Inversion
David Bechtoldt, Sidney Bender
cs.LG · cs.AI
Graph Neural Networks (GNNs) achieve strong predictive performance on graph-structured data across domains such as chemistry, biology, and network analysis, yet they provide no intrinsic explanation of their predictions. This limits their adoption in high-stakes and safety-critical settings. Counterfactual explanations address this by revealing the minimal structural modifications that would change a model's prediction. On graphs, however,...
-
16
Towards Truly Unsupervised Evaluation of Feature Selection
Hafiz Saud Arshad, Muhammad Rajabinasab, Arthur Zimek
cs.LG
Feature selection is one of the most important and fundamental tasks in data mining, tackled by a family of methods with an established set of evaluation techniques to measure the quality of a specific method. Most of the methods commonly used for the unsupervised evaluation of feature selection algorithms suffer from critical design flaws which question their unsupervised nature. In this paper, we provide a critical discussion on the...
-
17
Clustered Randomized Smoothing for Stochastic Prediction Functions
Eduardo Figueiredo, Frederik Mathiesen, Julian Schumann, Jens Kober, Arkady Zgonnikov, Luca Laurenti
cs.LG · eess.SY
Modern stochastic predictors can model rich, multi-modal outcome distributions. However, this expressive power comes with challenges in ensuring robust predictions $-$ a critical requirement in safety-critical domains. Randomized smoothing is a leading technique for improving robustness, particularly against adversarial perturbations. Yet, in stochastic multi-modal regression settings, randomized smoothing often fails due to mode collapse,...
-
18
Uncertainty-Aware Probabilistic Constrained Clustering from Entangled Pairwise Supervision
Shaojie Zhang, Ke Chen
cs.LG · cs.AI
Pairwise constrained clustering typically relies on hard must-link/cannot-link labels, whereas realistic pairwise supervision may be real-valued and entangle intrinsic ambiguity, expert judgment, and stochastic corruption. Existing deep constrained clustering (DCC) methods mainly target hard, expert-agnostic constraints, treating soft labels mostly numerically rather than semantically. We formalize this setting as uncertainty-aware...
-
19
SoftWater: Class-Aware Rate Allocation for Softmax Quantization
Joao V. Cavalcanti, Ashia C. Wilson
cs.LG
Post-training quantization pipelines routinely leave the softmax output layer in high precision. Yet in small LLMs with modern vocabularies, the head holds 15--30\% of all parameters, so a nominal ``2-bit'' model with an fp16 head can store several times as many bits per weight. We pose softmax-layer quantization as a rate-distortion problem under the KL divergence between the original and quantized output distributions. A second-order...
-
20
Reducing Symmetry Increase in Equivariant Neural Networks
Ning Lin, Jiacheng Cen, Anyi Li, Wenbing Huang, Hao Sun
cs.LG
Equivariant Neural Networks (ENNs) have empowered numerous applications in scientific fields. Despite their remarkable capacity for representing geometric structures, ENNs suffer from degraded expressivity when processing symmetric inputs: the output representations are invariant to transformations that extend beyond the input's symmetries. The mathematical essence of this phenomenon is that a symmetric input, after being processed by an...
-
21
Dual-Model Sentiment Analysis of Consumer Reviews in the Retail Coffee Sector Using Machine Learning and Deep Learning Approaches
Muntasir Hasan Kanchan, Md. Alamgir Hossain, Md. Samiul Islam, Muhammad Masud Tarek
cs.LG · cs.AI · cs.CV
Consumer reviews play an important role in shaping brand perception and business strategies, particularly in service-driven industries such as retail coffee. This study presents a comparative sentiment analysis framework for Starbucks customer reviews using classical machine learning and deep learning approaches. The dataset, collected from ConsumerAffairs, contains more than 700 reviews and was analyzed through preprocessing and exploratory...
-
22
Remote Sensing and Machine Learning-Based Analysis of Land Use and Vegetation Change in Dhaka District, Bangladesh
Muhammad Masud Tarek, Md. Alamgir Hossain, Md. Samiul Islam, Muntasir Hasan Kanchan
cs.LG · cs.AI · cs.CV
Rapid urbanization in Dhaka District, Bangladesh has triggered substantial alterations in land use and environmental conditions, necessitating systematic monitoring for informed urban planning and ecological sustainability. This study employs remote sensing data and machine learning techniques to analyze spatiotemporal changes in land cover and vegetation dynamics between 2019 and 2024. High-resolution satellite imagery from Sentinel-2 MSI...
-
23
TESLA: Taylor Expansion of Sinusoidal Learnable Activations
Daehwa Ko, Jaehyeon Kim, Seunghyun Ham, Jay Hoon Jung
cs.LG
The parity problem--deciding whether the number of ones in a binary vector is odd or even--remains challenging for standard neural networks due to linear inseparability and the need for global interactions. We propose TESLA, an activation defined as a learnable combination of sine and cosine terms, enabling explicit control over polynomial degree and selective amplification of high-order components. Theoretically, we show that constraining...
-
24
LoongReflect: Boosting Long-Horizon Reflection in Search Agents via Global Perspective Distillation
Zhixin Zhang, Xinke Jiang, Zhibang Yang, Weixuan Xu, Guohong Qiu, Xu Chu, Junfeng Zhao, Yasha Wang
cs.LG · cs.AI
Large language model agents increasingly rely on long-horizon reasoning to solve complex tasks involving planning, tool use, and memory. A critical capability in such settings is reflection: assessing trajectory progress, identifying missing evidence and unreliable intermediate states, and deciding whether to continue, revise, or abandon the current branch. Learning effective reflection, however, is challenging because reflection is performed...
-
25
TailBooster: A Dual-Layer Generative Framework for Extreme Value Augmentation with Operational Validity Enforcement
Karim Aly, Alexei Sharpanskykh, Jacco Hoekstra
cs.LG · cs.AI
Extreme events in air transport, such as severe arrival delays and abnormal air times, cause cascading network disruptions with substantial operational, economic, and safety costs. Such events are rare in historical records, leaving insufficient training signal for machine learning models. Synthetic data augmentation offers a principled solution, but conventional generative models under-represent distributional tails and give no guarantee...
-
26
Distillation of Foundation Models for Time-dependent PDEs
Daniel Musekamp, Boshra Ariguib, Andrei Manolache, Mathias Niepert
cs.LG
Foundation models for time-dependent partial differential equations (PDEs) are trained on large and diverse collections of physical systems and can generalize effectively to new downstream tasks. After fine-tuning on only a few trajectories from a target domain, they can achieve strong accuracy in low-data regimes. However, these models are typically large and computationally intensive, limiting their usefulness as fast surrogates for...
-
27
A Factor Graph Approach to Scalable Multi-Output Gaussian Process Regression
Wouter W. L. Nuijten, Esther G. van Pelt, Albert Podusenko, İsmail Şenöz, Wouter M. Kouw
cs.LG
Multi-output Gaussian process regression scales cubically in the number of observations times outputs, and dense kernel-matrix methods need bespoke handling whenever different outputs are observed at different inputs. We express multi-output Gaussian process regression as a Forney-style factor graph in which a nearest-neighbor chain orders a fixed candidate set of $C$ inputs into a one-dimensional sequence. Along this chain, latent Matérn...
-
28
Disentangling the Expressivity of RoPE
Selim Jerad, Anej Svete, Jiaoda Li, Ryan Cotterell
cs.LG · cs.FL
Two accounts recur in explanations of the success of rotary position embeddings (RoPE). Expressivity studies associate periodic position information with modular predicates, whereas mechanistic and long-context studies emphasize positional anchors and local offsets. We formalize both accounts for fully uniform, finite-precision soft-attention transformers. We find that, if every rotary component is periodic, RoPE transformers recognize...
-
29
DCM Bandits: Multiplayer Information Asymmetric Cascading Bandits for Multiple Clicks
Andy Wang, Charlton Shih, William Chang
cs.LG
In this work, we extend the Dependent Click Model (DCM) Bandits to a multiplayer information-asymmetric setting, where multiple agents interact with a shared ranked list and may observe multiple clicks per session, introducing new challenges for selection strategies. We study asymmetry in (1) actions and (2) rewards, providing sublinear regret guarantees for three settings where at least one asymmetry is present. Establishing matching...
-
30
Forward and Inverse Virtual Metrology for Phototransistor Gain: A Hierarchical, Uncertainty-Aware Approach for Small Production Datasets
Mahshid Amirabgir, Lorenza Ferrario, Paolo Conci, Mahdieh Amirabgir, Giancarlo Orengo
cs.LG · cs.CE · eess.SY
The customization, optimization and stabilization of the process flow of a silicon bipolar phototransistor commits months of cleanroom time before a finished device can be measured, so a model that predicts device gain from process parameters before a run has value out of proportion to its accuracy. We study this problem on a real fabrication history, thirteen to fourteen process runs of a single device: a small-sample, hierarchically...
-
31
Small-Scale Experiments: Are We There Yet?
Nicholas Lourie, Kyunghyun Cho, Karen Ullrich, Sanae Lotfi
cs.LG
Scaling laws promised cost-effective experiments; six years later, they have yet to fully deliver. Instead, researchers have found them unreliable at small scales (starting at 4M parameters) and concluded that sizable models cannot be avoided. We show this is not the case: the confounding factor is hyperparameters. Small models are highly sensitive, but hyperparameter sensitivity fades with scale. This small-scale sensitivity makes scaling...
-
32
Air Quality Station Simulation via LSTM and Attention-Based Modelling
Alexander Kostadinov, Petar O. Hristov, Dessislava Petrova-Antonova
cs.LG
Poor air quality in urban areas is driven by a complex chain of processes and presents a significant public health concern. To better understand and control the mechanisms that determine air quality, cities deploy networks of measurement stations, and launch initiatives for collecting denser data about the concentration of pollutants in the atmosphere. Extracting insights from the stations relies on their reliable and uninterrupted operation....
-
33
Kernel Methods for Learning Operators with Multiple Inputs and Outputs
Adrien Weihs, Chunyang Liao, Jingmin Sun, Hayden Schaeffer
cs.LG · math.ST · stat.ML
Learning mappings between infinite-dimensional objects is a central challenge in scientific machine learning. We introduce a general kernel-based encoder-decoder framework for operator learning that separates observation, representation, learning, and reconstruction. We develop this framework for multi-input, multi-output operator learning, where operators map between products of potentially distinct function spaces. Our approximation theory...
-
34
Towards Understanding On-Policy Distillation through the Lens of Test-Time Scaling
Xinmu Ge, Zizhuo Zhang, Yu Huang, Jianing Zhu, Lin Yuan, Wanli Gu, Weichang Wu, Weiran Huang, Xiaolu Zhang, Bo Han,...
cs.LG · cs.CL
On-policy distillation (OPD) has emerged as a promising post-training technique for enhancing LLM reasoning. It is commonly believed to enable the student model to distill knowledge from a stronger teacher model, thereby expanding capabilities beyond the pre-OPD base model. In this study, we examine this view through the lens of test-time scaling by varying the sampling budget K and evaluating performance with pass@K and avg@K. Specifically,...
-
35
Learning with Bilevel-Minimax Optimization for Efficient and Reliable Transfer Attacks
Yaohua Liu, Yifan Guo, Jiaxin Gao
cs.LG · cs.CV
Transfer-based adversarial attacks craft adversarial examples using surrogate models to mislead black-box victim models. Beyond perturbation generation, transferability is fundamentally governed by the coupling of initialization, surrogate adaptation, and gradient dynamics. We revisit this challenge from a bilevel-minimax perspective and propose BMAT (Bilevel-Minimax Adversarial Transfer). The bilevel formulation captures the dependency...
-
36
JAPE: Joint Anomaly Prediction and Intrinsic Explanation in Multivariate Time Series
Yian Wei, Yuanyuan Yao, Lu Chen, Xiangmin Zhou, Tianyi Li
cs.LG
Multivariate time-series anomaly prediction aims to identify whether and when anomalies will occur over a future horizon from historical observations. Existing methods primarily characterize anomalies as deviations in future numerical values, which may overlook subtle dependency changes induced by weak anomaly precursors and provide no native variable-level explanation together with the alert. To bridge these gaps, we propose JAPE, a Joint...
-
37
Orientation, not magnitude: the causal structure of task-vector interference in merged language models
Chencheng Zhu
cs.LG
Model merging by task arithmetic works until it doesn't, and the field diagnoses why with magnitudes: layerwise representation bias, deviations from cross-task linearity, parameter overlap. Tracking the exact layerwise cross-term of merged LLMs through a factorial ledger and intervening on it directly, we find magnitude insufficient - and inconsistent across model families - as a diagnostic axis. An exact decomposition of the layerwise flux...
-
38
High-Order Liquid Evidence Encoding for Gradual GNSS Spoofing Detection in Autonomous Driving
Muhammad Ayub Sabir, Junbiao Pang, Fatima Ashraf
cs.LG
Accurate Global Navigation Satellite System (GNSS)-based localization is essential for safe and reliable autonomous driving. However, spoofing attacks can manipulate vehicle position estimates. Continuous and subtle attacks are particularly difficult to detect because individual GNSS observations may remain plausible while the inconsistency between GNSS-implied displacement and onboard vehicle motion gradually increases. Existing methods...
-
39
TradingMoE: Routing the Right Experts in Evolving Markets
Chang Zhou, Xingtong Yu, Minbin Huang, Zhennan Wu, Yuan Fang, Hong Cheng, Xinming Zhang
cs.LG
Large language models (LLMs) have shown strong potential for financial analysis and trading, but direct trading remains challenging because the predictive capabilities required can vary across assets, decision fields, and market conditions. Existing LLM-based trading systems either coordinate human-defined external experts or adopt conventional internal Mixture-of-Experts (MoE) routers that do not directly evaluate how individual experts...
-
40
MOON: Multi-Objective OrthoNormalized Updates for Multitask Learning
Shiji Zhou, Kunlin Lyu, Lei Zhang, Ruodong Wang, Yifan Sun
cs.LG · cs.AI · stat.ML
Multi-objective optimization (MOO) has demonstrated significant success in multi-task learning by mitigating task conflicts through gradient manipulation. However, most existing methods flatten model parameters into vectors and perform gradient manipulation under Euclidean geometry, thereby overlooking the matrix structure prevalent in modern architectures such as Transformers. In this paper, we show that gradient manipulation in Euclidean...
-
41
Epiplexity Guided Data Selection and Generation for Out-of-Distribution Generalization
Ellen Su, Andres Potapczynski, Shikai Qiu, Edward Hughes, Andrew Gordon Wilson
cs.LG · cs.CL
Modern systems are increasingly expected to transfer across tasks not specified during training. What data facilitates generalization in these new, unanticipated settings? One hypothesis is that data with more structural information could contain shared circuits and subprograms that could be recycled in a wider array of downstream settings. Epiplexity, a recently proposed measure of the structural information a compute-bounded learner can...
-
42
Chain-of-Thought Shows the Path to a Tree: Realizing Branching Complexity
Debanjan Dutta, Anish Chakrabarty, Swagatam Das
cs.LG
Chain of Thought (CoT) lifts the expressive ceiling of bounded-depth Transformers, with characterizations tying the number of CoT steps to circuit complexity classes. What remains largely missing are concrete instantiations with explicit, depth-bounded constructions, and the traversal procedures such characterizations presuppose. We close this gap for branching complexity. We give CoT realizations of depth-first search (DFS) and of Dijkstra...
-
43
High-dimensional Multi-objective Bayesian Optimization with Learned Variable Interactions
Hongyan Wang, Jiayu Huang, Haotian Zheng, Xin Gao, Chi Ding, Ying Liu, Xia Wang, Qing Xu, Keqiang Li
cs.LG · cs.AI
Multi-objective Bayesian optimization (MOBO) is effective in identifying the Pareto fronts for expensive black-box problems. However, most current MOBO approaches are limited to low-dimensional decision space due to its exponential sampling complexity. This paper presents decision variable interaction analysis-based MOBO, ViaMOBO, a generic framework for expensive multi-objective problems with high-dimensional decision space. The key idea of...
-
44
Robust and Efficient Noisy-Label Time-Series Classification via Dynamic Time Warping Based Granular Ball Computing
Ziqiang Li, Yun Liu, Gouhei Tanaka
cs.LG · cs.AI
Dynamic Time Warping (DTW)-based Nearest-Neighbor (NN) classifiers are effective for time-series classification but are vulnerable to mislabeled training samples and require numerous DTW computations during inference. We propose DTW-based Granular Ball Computing (DTW-GBC), which organizes temporally similar training samples into granular balls and performs classification at the granule level. We further develop two granular-ball construction...
-
45
Consolidator: Learning Persistent Routed Memory Across Context Boundaries
Sungwoo Goo, Hwi-yeol Yun, Sangkeun Jung
cs.LG · cs.AI
Copying short-term memory (STM) into a slower store can preserve state across a context boundary, but persistence alone does not ensure that the retained state influences subsequent memory access. We test this distinction in a Phasor Memory Network (PMNet) using Consolidator, a shared slot-local operator that transforms routed STM before accumulating it into long-term memory (LTM), without replaying the source tokens. After each...
-
46
REOPD: Reliability-Adaptive Reward Extrapolation for On-Policy Distillation
Yang Sun, Lichao Ma, Houyuan Qin, Yuxin Liu, Hanyang Lu, Yao Zhu, Pinlong Cai, Guohang Yan
cs.LG · cs.AI
On-policy distillation (OPD) trains a student on its own trajectories under dense token-level supervision from a teacher. Reward-extrapolation methods such as ExOPD amplify the teacher-reference log-likelihood ratio to move beyond direct imitation, but apply a single global coefficient $λ$ to every token. This can drive the student to fit extreme peaks in the implicit reward, causing reward hacking and unstable training, and the optimal $λ$...
-
47
LEMUR: Latent Entropy-aware Multimodal Unlearning via Visual-anchored Reasoning Redirection
Xinhao Zhong, Yuxia Qiao, Junhao Li, Hao Fang, Yi Sun, Bin Chen
cs.LG · cs.CL
Reinforcement-learning (RL) post-training equips multimodal large reasoning models (MLRMs) with exploratory chains of thought (CoT), substantially improving visual reasoning. However, we find that this capability introduces a distinct privacy vulnerability: even when a sensitive fact is successfully unlearned from the final answer, the model may still reproduce it in its reasoning trace. This leakage is substantially more pronounced in...
-
48
Drift and Dependence: Layer-wise Information-Theoretic Bounds for Replay-Based Continual Learning
Tieliang Gong, Zhongbo Zhang, Wen Wen, Yong-Jin Liu
cs.LG · stat.ML
Continual learning must absorb new tasks without erasing old ones, and replay---mixing a small buffer of past examples into current training---is among the most effective remedies for catastrophic forgetting. Yet its generalization behavior is shaped by two coupled effects that existing analyses fold into a single hypothesis-level quantity: finite memory replaces each past distribution with an empirical proxy, and repeated reuse couples the...
-
49
FunnelCausalNet: Funnel-aware Joint Conversion-Revenue Uplift for Multi-tier Coupon Allocation
Yu Zhang, Zhihan Wang, Guanlin Chen, Min Jiang, Shuai Li
cs.LG · cs.IR
Coupon campaigns seek to lift both conversion and revenue, but gross merchandise value (GMV) follows a deterministic funnel from conversion to conditional order value and is zero-inflated and heavy-tailed. We propose FunnelCausalNet, an uplift estimator coupling a binary conversion head with a nonnegative conditional-value head through $μ_{\mathrm{gmv}}=μ_{\mathrm{conv}}μ_{\mathrm{val}}$. Under explicit RCT, support, rate-gap, and cross-head...
-
50
GCPO: Diagnosing and Constraining Subspace Geometry in Rollout RL for LLMs
Kai Yang, Jingwei Xu, Wanyu Wang, Kai-Yuan Guo, Zhenbo Yu, Yi Wang, Yu Qiao
cs.LG · cs.AI
On-policy rollout methods such as GRPO are central to post-training of large language models, yet they frequently suffer from training instabilities, cross-task capability degradation, and response-length inflation. Although prior work has characterized the subspace geometry of aggregate updates, the stepwise variation of this geometry and its relationship to model performance remain unclear. We introduce Principal-Subspace Overlap, a...
-
51
Rubric Dropout: A Simple Way to Mitigate Reward Hacking in Rubric-as-Reward RL
Minglai Yang, Xinyu Guo, Utkarsh Tyagi, Mian Zhang, Razvan Dumitru, Sunjie Hou, Yunzhong He, Daniel Yue Zhang, Ying Liu
cs.LG · cs.AI · cs.CL
Reinforcement learning against rubrics, lists of criteria graded by an LLM judge, has become a standard way to post-train language models on tasks with no deterministic answer. The rubric, however, is a fixed proxy for quality, never a complete description of it, and a policy trained against it long enough will learn to exploit the difference. We measure this directly. Training Qwen3-8B with Group Relative Policy Optimization (GRPO) on...
-
52
Low-Interaction-Rank Learning: Unifying Multiplicative Dual-Encoder Heads
Zijian Zhao, Sen Li
cs.LG · cs.AI
A multiplicative dual-encoder network computes a real-valued output for a pair of inputs as the inner product of their separate encodings. This architecture has been developed independently in operator learning, bipartite matching, contrastive vision-language models, retrieval, and other areas, yet no unified theory guides the basic design decisions: how many interaction modes to represent, how to normalize the encoders, and when the...
-
53
Is Per-Agent Policy Composition Safe? Rethinking Successor-Feature Transfer in Cooperative Multi-Agent Reinforcement Learning
Zijian Zhao, Sen Li
cs.LG · cs.AI · cs.MA
Many reinforcement learning systems, from fleet management to traffic signal control, must serve an objective that changes dynamically after deployment, and retraining a policy for each new objective is prohibitively expensive. For a single agent, this problem is well understood: successor features with generalized policy improvement, together with their universal extension, recombine a library of learned policies into a policy for any new...
-
54
Continuous-Latent Predictive Modeling with Semantic Alignment for EEG-Language Foundation Models
Myeong-Ju Cho, Hye-Bin Shin, Seo-Hyun Lee, Seong-Whan Lee
cs.LG
Recent advances in EEG foundation models have demonstrated the potential of large-scale pretraining to enable generalizable neural decoding across subjects, recording environments, and datasets. However, dominant pretraining paradigms face key challenges: masked autoencoding tends to prioritize low-level signal reconstruction over task-relevant semantics, while autoregressive modeling creates a mismatch between continuous neural dynamics and...
-
55
Towards a Formal Definition of Agent Memory: Basis, Span, Optimality, and the Sequential Memory Problem
Hongyao Tang
cs.LG
Despite the wide deployment of memory in large-model agents, there is no unified formal account of what a memory is or when it is optimal. This paper takes a first step toward this account. The central idea is that memory is a basis, knowledge is its span, and answerability is a coverage problem: an agent stores events extracted from a material; a generation operator turns any event set into the knowledge it entails; and a query is answerable...
-
56
Transferable Above-Ground Biomass (AGB) Estimation Model from Multi-Sensor Data with Sparse Field Calibration
Pann Thinzar Seint, Bryan Atwood, Subas Chhatkuli
cs.LG · cs.CV
Spatially continuous quantification of forest above-ground biomass (AGB) is what makes carbon accounting credible and mitigation strategies actionable. While field inventories provide high localized accuracy, they are spatially sparse; conversely, spaceborne LiDAR from the Global Ecosystem Dynamics Investigation (GEDI) offers broad biomass samples but lacks spatial continuity and systematic underestimation of high-biomass forests. This paper...
-
57
FM-LLM: A frequency-enhanced mixture-of-experts framework for adapting LLMs to time series forecasting
Rentao Gu, Yihang Ding, Junjie Li, Yi Ding, Weijing Sang, Xiaoli Huo, Xin Qin, Yuefeng Ji
cs.LG · cs.AI · cs.NI · eess.SP
Recent advances in Large Language Models (LLMs) have spurred cross-modal solutions for time-series forecasting. However, existing methods rely heavily on textual prompts for modality alignment-introducing nontrivial computational overhead and failing to leverage the rich spectral dynamics inherent in time-series data. To enable prompt-free, frequency-aware adaptation of frozen LLMs, we propose FM-LLM (Frequency-Enhanced Mixture-of-Experts for...
-
58
A Local Sinkhorn Framework for Conditional Distribution Reconstruction of Multidimensional Random Fields
Mingtao Xia, Qijing Shen
cs.LG
In this paper, we propose a local Sinkhorn divergence framework for conditional distribution reconstruction of multidimensional random fields. By utilizing the debiased Sinkhorn divergence, our proposed approach develops a differentiable and computationally efficient local distribution matching objective to train stochastic neural networks (SNNs). Furthermore, we establish theoretical generalization error estimates for our local Sinkhorn...
-
59
Dion3: Full-Stack Orthogonal Updates
Noah Amsel, Jack Zhang, Kwangjun Ahn, Ali Naeimi, Austin Feng, Berlin Chen, Tri Dao, John Langford
cs.LG · cs.AI
The Muon optimizer incurs a significant overhead cost due to its cubic-time Newton-Schulz orthogonalization step. When weights are sharded, communication overhead compounds this computational cost, eroding the benefits of Muon in many settings. We present Dion3, a revision of Muon that targets this overhead at every level of the stack. Our Gram Newton-Schulz algorithm reduces the FLOP cost of orthogonalization, our CuteDSL kernels accelerate...
-
60
RECAST: A Machine-Learning Framework for Correction and Super-Resolution of Coarse-Grid PDE Solvers
Maryam Reza, Farbod Faraji
cs.LG · physics.comp-ph
Coarse-grid numerical solvers can substantially reduce the computational cost of time-dependent PDE simulation, but under-resolution often degrades both the trajectory and the spatial fidelity of the solution. We introduce RECAST (Recurrent Error Correction And Super-resolution of coarse-grid Trajectories), a machine-learning framework designed to restore this lost accuracy while retaining coarse-grid evolution. RECAST combines learned...
-
61
Sparse and robust geometric twin support vector machine via asymmetric RoBoSS loss function
Kai Qi, Xinji Huang, Hongchun Wang
cs.LG
In real-world scenarios, the training data usually contains redundant features, label noise and feature noise, which provide severe challenges for the efficiency of machine learning methods. Since standard support vector machine (SVM) adopts $l_2$-norm penalty and hinge loss function, it lacks the ability of selecting significant features and is sensitive to noise. To address these issues, this paper proposes a novel asymmetric, robust,...
-
62
When Offline Evaluation Misleads: A Diagnostic Protocol for Reward and Policy Selection in Delayed-Feedback Contextual Bandits
Sang Su Lee, Vineeth Loganathan, Shishir Dash, Vijay Raghavan
cs.LG
Personalizing marketing messages with contextual multi-armed bandits (CMABs) drives real business value, yet the objective that ultimately matters - a downstream conversion - is observed only weeks later, too late to drive online learning. Teams therefore train the bandit on a fast proxy reward, and separately must judge whether a contextual bandit is worth its complexity over sending one best message. Settling both decisions with the usual...
-
63
Certifying What Helps Customer-Return Timing: A Screen-and-Confirm Test for Conditioning Signals, and Why Decay Is Nearly Enough
Sang Su Lee, Vineeth Loganathan, Shishir Dash, Vijay Raghavan
cs.LG · stat.AP
Practitioners enrich customer-return models with ever more signals (lifetime value, category, recency/frequency, calendar, geography), and the temporal-point-process (TPP) literature follows suit with covariate- and external-covariate-conditioned intensities. But does any of it improve the timing, and how would you know? A null ("feature X doesn't help") is only meaningful if the model could have found a signal. We make two contributions--a...
-
64
Robust Ambiguity Detection (RAD) From Model- and Feature-Space Consistency
Manya Singh, Mark T. Keane, Arjun Pakrashi
cs.LG
Machine learning models should be robust, in the sense of remaining predictively consistent under permissible variations. A model's predictions should ideally remain unchanged when it is replaced by a functionally equivalent one, or when its inputs are subject to minor, admissible perturbations. If such changes alter a prediction significantly, then the prediction is "ambiguous" with respect to the model. Models should abstain from making...
-
65
Hierarchical Federated Transfer Learning in Digital Twin-Based Vehicular Networks
Qasim Zia, Saide Zhu, Haoxin Wang, Zafar Iqbal, Yingshu Li
cs.LG · cs.AI
In recent research on the Digital Twin-based Vehicular Ad hoc Network(DT-VANET), Federated Learning (FL) has shown its ability to provide data privacy. However, Federated learning struggles to adequately train a global model when confronted with data heterogeneity and data sparsity among vehicles, which ensure suboptimal accuracy in making precise predictions for different vehicle types. To address these challenges, this paper combines...
This edition is part of The Daily Abstract — cs.LG archive. Subscribe to receive these in your inbox each morning, automatically translated to Spanish, with reply-to-PDF: arxivdaily.ignorelist.com.
#D99C5E. Built and served on an always-free VM. The masthead is set 14% letterspaced because newspapers do that and it works.