cs.LG · 2026-07-14 · No. 53

Machine Learning, 2026-07-14.

59 new papers in cs.LG. Titles, authors, abstracts. Links to arXiv. Want this in your inbox every morning? Subscribe →

01 — The papers

59 entries
  1. 01

    Requential Coding: Pushing the Limits of Model Compression with Self-Generated Training Data

    Shikai Qiu, Marc Finzi, Yujia Zheng, Kun Zhang, Andrew Gordon Wilson

    cs.LG

    Compression is fundamental to intelligence. A model that can represent its training data as a short code has discovered regularities that enable generalization. Large neural networks may learn functions far simpler than their parameter counts suggest, but it is challenging to construct codes that realize this simplicity. Parameter-based methods such as quantization produce code lengths that scale with model size, insensitive to how much...

    arxiv.org/abs/2607.11883 · PDF

  2. 02

    Invariant Learning Dynamics of Transformers in Inductive Reasoning Tasks

    Tiberiu Musat, Tiago Pimentel, Nicholas Zucchet, Thomas Hofmann

    cs.LG · cs.AI

    We present a theoretical framework to explain the emergence of inductive reasoning abilities in Transformer language models. While previous works on Transformer learning dynamics have so far been mostly tied to specific tasks, we study a generalized class of inductive tasks that unifies several synthetic tasks known in the literature, including in-context n-grams and multi-hop reasoning. In this class, we theoretically prove that the training...

    arxiv.org/abs/2607.11875 · PDF

  3. 03

    Inside the Unfair Judge: A Mechanistic Interpretability Account of LLM-as-Judge Bias

    Zixiang Xu, Sixian Li, Huaxing Liu, Xiang Wang, Shuai Li, Zirui Song, Xiuying Chen

    cs.LG · cs.AI · cs.CL

    Existing studies of LLM-as-judge scoring bias work predominantly at the input-output level: they perturb inputs, measure score deltas, and propose prompt-level mitigations. We argue that the same biases admit a representation-level account in the judge's hidden state, complementary to the input-output view and operationally useful in ways it does not afford. We report three findings, across seven judges, seven bias types, and nine benchmarks....

    arxiv.org/abs/2607.11871 · PDF

  4. 04

    Transformer-Guided Swarm Intelligence for Frugal Neural Architecture Search

    Romain Amigon

    cs.LG · cs.AI · cs.NE

    Neural Architecture Search (NAS) has automated the design of deep learning models but traditionally requires massive computational resources, often measured in thousands of GPU-days. In this paper, we propose a frugal and memetic NAS framework designed to democratize architecture design on consumer-grade hardware. Our approach combines the global macro-search capabilities of an autoregressive Transformer controller, trained via Reinforcement...

    arxiv.org/abs/2607.11826 · PDF

  5. 05

    Relaxing Faithfulness with Intervention-Only Causal Discovery

    Bijan Mazaheri, Jiaqi Zhang, Caroline Uhler

    cs.LG · stat.ML

    Causal discovery algorithms learn a network that describes the causal dependencies among random variables. A common workflow involves first utilizing conditional independence properties on observational data to determine partially directed causal relationships, then applying interventions to orient the unknown causal directions. A critical assumption for the first step is faithfulness: a requirement that causally linked variables exhibit...

    arxiv.org/abs/2607.11816 · PDF

  6. 06

    An Exact Instrument for State Usage in Selective State-Space Models, and the Input-Driven Migration It Reveals

    Raktim Bhattacharya

    cs.LG

    Selective state-space models such as Mamba route information through a bank of first-order modes whose input coupling is set by a learned selection mechanism. We give an exact instrument for measuring how a trained model uses these modes. Because the state matrix is diagonal, each channel's output decomposes exactly into per-mode contributions, and a per-(layer, channel, window) Gram tensor yields the exact output error of dropping any subset...

    arxiv.org/abs/2607.11796 · PDF

  7. 07

    From Expressivity to Sample Complexity: Narrow Teachers for Transformers via C-RASP

    Michael Rizvi-Martel, Satwik Bhattamishra, Guillaume Rabusseau, Michael Hahn

    cs.LG · cs.CL

    A theoretical understanding of Transformers is crucial to better understand the capacities and limitations of large language models (LLMs). There is much work analyzing the expressivity of attention-based models. By proposing handcrafted weights or using computational complexity arguments, a large amount of past theoretical works have sought to characterize which tasks are and which are not in the hypothesis class of Transformer models....

    arxiv.org/abs/2607.11760 · PDF

  8. 08

    From Global to Factor-Wise Expert Composition in Discrete Diffusion Models

    Haozhe Huang, Yudong Xu, Abhijoy Mandal, Alán Aspuru-Guzik

    cs.LG

    Discrete diffusion models offer a powerful framework for solving complex reasoning tasks, particularly through compositional generation, which combines multiple pre-trained experts to generalize beyond their individual training data. Recent theoretical corrections introduce time-dependent mixing weights to better align composed diffusion dynamics with the intended target. However, these methods are fundamentally limited by working on a...

    arxiv.org/abs/2607.11758 · PDF

  9. 09

    HiFi-LLP: High-Fidelity, Low-Cost Latency Predictors with Confidence for Robust HW-NAS

    Shambhavi Balamuthu Sampath, Behzad Shomali, Nael Fasfous, Moritz Thoma, Judeson Anthony Fernando, Lukas...

    cs.LG · cs.AR

    With deep neural networks (DNNs) increasingly deployed on edge devices, hardware (HW)-aware optimization techniques--such as HW-aware compression and HW-aware neural architecture search (HW-NAS)--have become essential. These methods rely on real feedback from the target hardware to tailor DNN architectures for efficient deployment. While the search can be parallelized, latency measurements via hardware-in-the-loop (HIL) remain a bottleneck...

    arxiv.org/abs/2607.11746 · PDF

  10. 10

    Time-Lag-Aware Deep Reinforcement Learning for Flexible Job-Shop Scheduling in PPVC Module Factories

    Ziheng Zhang, Wei Zhang

    cs.LG · cs.AI · math.OC

    Prefabricated prefinished volumetric construction moves most building work into module factories, whose production floor operates as a flexible job shop. A major complication is decisive: long post-operation time-lags caused by concrete curing, watertightness ponding tests, and paint drying, during which a module is blocked while its workstation stays free. On benchmark instances grounded in an official national prefabrication guidebook,...

    arxiv.org/abs/2607.11725 · PDF

  11. 11

    Active Offline-to-Online Reinforcement Learning

    Alper Kamil Bozkurt, Shangtong Zhang, Yuichi Motai

    cs.LG · cs.AI

    Background: Offline reinforcement learning (RL) enables effective policies to be trained from large, previously collected datasets and subsequently improved through limited online interaction. This offline-to-online RL (O2O-RL) paradigm is particularly promising in nonstationary domains where interaction is costly or potentially hazardous. Standard O2O-RL pipelines train multiple candidate policies offline, evaluate them using off-policy or...

    arxiv.org/abs/2607.11720 · PDF

  12. 12

    CatRetriever: Contrastive Representation Learning for Slab-to-Bulk Retrieval in Generative Catalyst Discovery

    Jungho Oh, Woosung Kim, Dong Hyeon Mok, Jonggeol Na, Seoin Back

    cs.LG · cond-mat.mtrl-sci

    Inverse design is an emerging data-driven paradigm for efficiently navigating vast chemical spaces to discover new materials with targeted properties, and in the context of heterogeneous catalysis, surface generative models have recently advanced this goal by directly generating catalyst surface-adsorbate structures. However, these models typically operate at the slab level and do not provide the corresponding parent bulk structure, making it...

    arxiv.org/abs/2607.11712 · PDF

  13. 13

    A multi-scale feature enhanced graph neural network for fluid dynamics prediction in complex geometries

    Li Xiao, Tianyu Li, Yiye Zou, Mingjie Zhang, Xiaogangd Deng

    cs.LG · physics.flu-dyn

    Industrial design in fields such as vehicle and aerospace engineering often relies on large-scale numerical simulations to evaluate fluid dynamics performance, which can incur substantial computational costs. Deep neural networks have shown promise in improving simulation efficiency, especially graph neural networks (GNNs), which demonstrate great potential due to their flexibility with unstructured data. However, GNNs face challenges when...

    arxiv.org/abs/2607.11672 · PDF

  14. 14

    How to Tame Grokking: Representation Geometry as a Control Signal

    Maksim A Kazanskii

    cs.LG

    Grokking is a phenomenon in which neural networks initially memorize training data and only later exhibit strong generalization after prolonged optimization. Despite extensive recent study, the factors influencing the emergence and timing of grokking remain incompletely understood. We investigate the relationship between representation geometry and delayed generalization. We find that dimensionality collapse consistently precedes the onset of...

    arxiv.org/abs/2607.11666 · PDF

  15. 15

    Bet on Features: Anytime-Valid and Feature-Aware Auditing of Conditional Quantile Forecasters

    Ivane Antonov, Sohom Mukherjee, Richard Pibernik, Yo Joong Choe

    cs.LG · stat.ML

    Black-box conditional quantile forecasts are widely used for sequential decisions under asymmetric costs, such as inventory planning in supply chain management. Once deployed, such forecasters must be monitored continuously as data streams drift and regimes change; this invalidates standard, fixed-horizon backtests for calibration. Further, existing backtests do not take into account that the notion of calibration is, in fact,...

    arxiv.org/abs/2607.11653 · PDF

  16. 16

    Fundamental Limitations of Fixed-Budget Best-Arm Identification

    Motti Goldberger

    cs.LG · stat.ML

    In fixed-budget best-arm identification, also known as ranking and selection, an algorithm has a sampling budget to distribute across $K$ arms. Each sample provides noisy feedback about that arm's mean, and the goal is to identify the arm with the largest mean. A common performance benchmark is the static oracle: a non-adaptive strategy that knows the means in advance and chooses fixed sampling proportions to maximize the exponential decay...

    arxiv.org/abs/2607.11635 · PDF

  17. 17

    Privacy-Aware Collaborative and Distributed Bayesian Optimization

    Aditya Rane, Sathwik Yamana, Paritosh Ramanan, Srikanthan Ramesh, Akash Deep

    cs.LG · stat.ME

    We propose a collaborative meta-learning framework for distributed Bayesian optimization matching centralized performance without raw-data exchange. We show gradient sharing leaks client observations, with leakage worsening as the search converges and queries concentrate near the optimum. We evaluate a differentially private defense and characterize its privacy-utility trade-off.

    arxiv.org/abs/2607.11600 · PDF

  18. 18

    DiffEEG: A Self-Supervised Denoising Diffusion Model for Learning EEG Generic Representations

    Abdulkader Helwan, Lina Abou-Abbas, Hussein El Amouri, Belkacem Chikhaoui, Khadidja Henni

    cs.LG · cs.AI · cs.CV · eess.SP

    Deep learning for EEG-based seizure detection faces critical challenges: severe annotation scarcity and extreme class imbalance, where ictal events comprise less than 10\% of clinical recordings. We present DiffEEG, a 9.6M-parameter self-supervised foundation model that addresses both limitations through denoising diffusion pre-training and reinforcement learning (RL)-based fine-tuning. Pre-trained on 1.3M unlabeled segments from the Temple...

    arxiv.org/abs/2607.11578 · PDF

  19. 19

    Structure-Feature Aligned Graph Learning via Alternating Constrained Optimization

    Chengcheng Yan, Qingsong Wang

    cs.LG · cs.AI

    We introduce a constrained two-view framework for node prediction that aligns structure-conditioned GNN embeddings with a structure-free feature prior learned by an anchor model. Conventional Graph Neural Networks (GNNs) couple feature transformation and neighborhood aggregation, which renders them vulnerable to topology noise and heterophilous connections. To decouple this dependency, our framework utilizes an independent anchor network to...

    arxiv.org/abs/2607.11577 · PDF

  20. 20

    Heuristic Learning for Active Flow Control Using Coding Agents

    Paul Garnier, Jonathan Viquerat, Elie Hachem

    cs.LG · cs.AI · physics.flu-dyn

    Active flow control involves nonlinear dynamics, partial observations, and computationally expensive simulations, making controller design particularly challenging. Deep reinforcement learning (DRL) has emerged as a powerful framework for such problems, but its success typically relies on large numbers of simulator interactions and produces neural-network policies whose decision process often remains difficult to interpret. In this work, we...

    arxiv.org/abs/2607.11565 · PDF

  21. 21

    Advancing Optimal Subset Oracle via Learning Relaxation of Neural Set Functions

    Yongquan Shi, Zijing Ou, Shiping Wang, Yatao Bian

    cs.LG

    Learning neural set functions is pivotal to a wide range of important applications, including compound selection in AI-driven drug discovery and product recommendation. Recent work has introduced optimal subset oracles to implicitly learn set functions under practical weakly supervised settings, where model parameters are optimized through mean-field variational inference. However, these frameworks rely on Monte Carlo sampling to estimate...

    arxiv.org/abs/2607.11555 · PDF

  22. 22

    Condition-Stratified Robustness Analysis of Post-Hoc Calibration Methods for Probabilistic Classifiers

    Gurdeep Singh Virdee

    cs.LG

    Post-hoc calibration is widely adopted to correct probability estimates from trained classifiers, yet most evaluations report aggregate performance without testing whether that performance holds across distinct operating conditions within a single dataset. We present a pre-registered, condition-stratified robustness analysis comparing temperature scaling (TEMP) and isotonic regression (ISO) across four controlled conditions (C1--C4). Four...

    arxiv.org/abs/2607.11542 · PDF

  23. 23

    Random Label Prediction Heads for Studying Memorization in Deep Neural Networks

    Marlon Becker, Jonas Konrad, Luis Garcia Rodriguez, Benjamin Risse

    cs.LG

    We introduce a straightforward yet effective method to empirically study memorization in deep neural networks for classification tasks. Our approach augments each training sample with auxiliary random labels, which are then predicted by a random label prediction head (RLP-head). RLP-heads can be attached at arbitrary depths of a network, predicting random labels from the corresponding intermediate representation and thereby enabling analysis...

    arxiv.org/abs/2607.11541 · PDF

  24. 24

    AutoMatBench: An Automatic Optimization Toolkit for the Acceleration of Material Properties Prediction Benchmarking

    Hongxiao Li, Wanling Gao

    cs.LG · cs.AI

    Material property prediction (MPP) infers key properties from chemical composition and structure, accelerating the discovery and optimization of novel materials. In the realm of MPP, MatBench is a widely accepted benchmarking tool that defines over ten significant problems and provides the paradigm of performance evaluation for AI prediction models. Even though MatBench works well in benchmarking the performances of prediction models on...

    arxiv.org/abs/2607.11526 · PDF

  25. 25

    DAG-FM: A Foundation Model for Causal Discovery under Heterogeneous Causal Mechanisms

    Yikang Chen, Zhengkang Guan, Haoyuan Qian, Peng Cui, Yi Yang, Kun Kuang

    cs.LG · stat.ML

    Causal discovery from observational tabular data remains fundamentally challenging, primarily due to the heterogeneity of underlying causal mechanisms and the high-dimensional combinatorial search space of Directed Acyclic Graphs (DAGs). In this paper, we propose \textbf{DAG-FM}, a novel foundation model architecture that amortizes causal discovery. Unlike direct matrix prediction, DAG-FM decomposes the causal discovery process into two...

    arxiv.org/abs/2607.11510 · PDF

  26. 26

    CDFM: Towards a General-Purpose Causal Discovery Foundation Model

    Jie Qiao, Ruichu Cai, Zijian Li, Weilin Chen, Pengfei Hua, Boyan Xu, Zhengming Chen, Zhifeng Hao, Peng Cui

    cs.LG · cs.AI · stat.ML

    Causal discovery, the process of recovering underlying causal structures from observational data, is a fundamental pursuit across scientific disciplines. Over the past decades, numerous algorithms have been developed to tackle this challenge through workflows tailored to the specific causal mechanisms underlying each type of dataset, demonstrating effectiveness across a wide range of applications. However, as the volume and heterogeneity of...

    arxiv.org/abs/2607.11508 · PDF

  27. 27

    SCOPE-RL: Optimizing Reasoning Paths Before and After Success

    Xiaojian Liu, Han Xu, Jianqiang Xia, Zhixuan Li, Ke Xu, Yiwei Dai, Xinran Chen, Changwo Wu, Yuchen Li

    cs.LG · cs.CL

    Reinforcement learning with verifiable rewards (RLVR) optimizes LLMs using sparse verifiable final-answer rewards. This sparse anchor reliably verifies whether a trajectory succeeds but provides no direct feedback on the reasoning path that produced it. Before success, prerequisite progress on hard problems receives no reward signal; after success, outcome rewards cannot distinguish well-organized correct trajectories from redundant or...

    arxiv.org/abs/2607.11506 · PDF

  28. 28

    Proxy Exploration and Reusable Guidance: A Modular LLM Post-Training Paradigm via Proxy-Guided Update Signals

    Daocheng Fu, Rong Wu, Yu Yang, Xuemeng Yang, Jianbiao Mei, Licheng Wen, Pinlong Cai, Yong Liu, Botian Shi, Yu Qiao

    cs.LG · cs.AI

    Post-training is essential for refining the domain-specific capabilities of large language models (LLMs), yet existing reward optimization and distribution matching methods tightly couple policy exploration with distribution alignment. This coupling forces expensive exploration directly on the policy model and severely hinders the asynchronous generation, reuse, and cross-model transfer of optimization signals. In this paper, we propose...

    arxiv.org/abs/2607.11505 · PDF

  29. 29

    IG-GAN: A Generative Adversarial Network for Aerodynamic Data Generation Based on Intrinsic Geometry

    Ying Yan, Liwei Hu, Xiaoming Zhang

    cs.LG · cs.AI

    Existing generative models learn data distributions in flat Euclidean space. However, most data in our real world are manifolds embedded in high dimensional Euclidean space. Therefore, we propose an intrinsic-geometry-based generative adversarial network (IG-GAN) for data generation in the field of aerodynamics. The generator of the IG-GAN represents aerodynamic data as a piecewise smooth manifold constructed by Bézier surfaces, and the...

    arxiv.org/abs/2607.11497 · PDF

  30. 30

    Agentic Skill Optimization over Lie Algebroids

    Sridhar Mahadevan

    cs.LG · cs.AI · math.CT

    Agentic systems increasingly improve themselves by editing skills: prompts, rubrics, plans, tool contracts, examples, validators, and traces. Skill edits are not independent coordinates in a vector space: they are local repairs to structured artifacts whose effects are observed only after rollout, validation, and critique. Distinct edits can have the same immediate visible effect while differing in routing context, template state, guardrail...

    arxiv.org/abs/2607.11493 · PDF

  31. 31

    HyperSafe: Inference-Time Safety Recovery for Fine-Tuned Language Models

    Aznaur Aliev, Carlos Hinojosa, Abdelrahman Eldesokey, Bang An, Bernard Ghanem, Yibo Yang

    cs.LG · cs.CL

    Safety alignment in large language models can be fragile under fine-tuning, as even benign task adaptation may increase harmful compliance. Existing defenses mainly follow two directions: they either intervene during or after fine-tuning through retraining or weight modification, which can be costly and may hurt task performance, or they use model-agnostic safety classifiers, which may miss failures specific to a given fine-tuned checkpoint....

    arxiv.org/abs/2607.11475 · PDF

  32. 32

    Event-based Neural Decoding for Neuroprosthetic Motor Control

    Khaleelulla Khan Nazeer, Sirine Arfa, Matthias Jobst, Richard George, Christian Mayr

    cs.LG · cs.NE

    A substantial number of patients experience diminished mobility due to disabilities, diseases, or accidents. Although modern prostheses, powered by deep neural networks, hold the promise of significantly enhancing the quality of life for these individuals, their widespread adoption is hindered by significant latency, energy consumption, and spatial requirements. Wired connections to external high-performance processors restrict patient...

    arxiv.org/abs/2607.11445 · PDF

  33. 33

    Velocity Scheduled Flow Matching

    Vitalii Bondar

    cs.LG

    Flow matching trains a neural network to regress the conditional velocity along a linear interpolant between noise and data, and the number of network evaluations~(NFE) sets the cost of sampling. The straight-line interpolant carries an implicit choice: the sample moves at constant speed throughout the trajectory. We relax this choice and introduce Velocity Scheduled Flow Matching~(VSFM), which replaces the conditional target $x_1 - x_0$ with...

    arxiv.org/abs/2607.11442 · PDF

  34. 34

    Generalizing Preference-based Reinforcement Learning: a Rationality Model for Incomparability

    Simone Drago, Marco Mussi, Leonardo Bianconi, Alberto Maria Metelli

    cs.LG

    In this work, we study the reinforcement learning (RL) problem from pairwise trajectory comparisons provided by a human expert. We generalize preference-based RL by formalizing a novel setting in which the expert can also label trajectory pairs as incomparable, i.e., when neither trajectory dominates the other. We introduce the learning problem and the desiderata that its solution should satisfy. Then, we propose a novel...

    arxiv.org/abs/2607.11432 · PDF

  35. 35

    Physics-Aware Conditional SetGAN for Spatially Consistent Multi-User TR 38.901 Channel Generation

    Mauro Gonzalo Tarazona-Levano, David Lopez-Perez, Nicola Piovesan, David Gomez-Barquero

    cs.LG

    TR 38.901-based channel models such as Sionna are reliable, but generating many multi-user channel realizations remains expensive. This paper asks a practical question: can a trained generative model produce multi-user TR 38.901 channels faster than Sionna without losing the spatial correlations imposed by user geometry? To answer this question, we propose a physics-aware, geometry-conditioned SetGAN trained on Sionna reference data. The...

    arxiv.org/abs/2607.11429 · PDF

  36. 36

    Surprisingly Simple and Effective Multi-Domain Graph Foundation Model through Graph-to-Table Alignment

    Chunyu Hu, Tianyin Liao, Ge Lan, Xingxuan Zhang, Jianxin Li, Peng Cui, Ziwei Zhang

    cs.LG

    Graph Foundation Models (GFMs) have emerged as a promising paradigm for learning transferable representations across diverse graph domains. Recent advancements in GFMs have been largely dominated by two paradigms: Graph Neural Network and Large Language Model (LLM) based methods. However, these methods often face a fundamental dilemma between training with limited data and a heavy reliance on textual attributes. Tabular foundation models...

    arxiv.org/abs/2607.11374 · PDF

  37. 37

    BackgroundMellow: A Multi-Modal Cohesive Framework for Narrative-Driven Rich Cinematic Soundscape Generation

    Ajitesh Jamulkar, Aritra Hazra

    cs.LG · cs.AI · cs.MM

    Generating immersive, synchronized and cinematic audio for long-form textual narratives remains a significant challenge in multi-modal AI. While current Text-to-Audio (TTA) frameworks successfully synthesize isolated sound effects, they struggle with narrative cohesion, temporal alignment, and cinematic emotional depth. We present BackgroundMellow, a framework that treats story-to-audio generation as a precise orchestration and signal...

    arxiv.org/abs/2607.11364 · PDF

  38. 38

    PRISM Edit: One Vector for All Temporal Answers

    Chen Huang, Qi Zheng, Ruiqin Zheng, Long Zeng, Yuantong Xu

    cs.LG · cs.AI

    Model editing keeps large language models (LLMs) up to date without retraining, but temporal facts expose a limitation of the prevailing locate-and-edit paradigm: an update is not always a replacement. When a fact changes, the new answer should become current while the old answer may remain correct in historical time contexts. Building on this insight, we use causal tracing to show that LLMs already support this distinction via a two-stage...

    arxiv.org/abs/2607.11327 · PDF

  39. 39

    SPARC-Net: A Spectral, Causality-Aware, and Hard-Constrained Physics-Informed Architecture for Stiff and Shock-Dominated Partial Differential Equations

    Divyavardhan Singh, Dimple Sonone, Hammad Mohammad, Kishor Upla

    cs.LG · cs.DM

    Physics-Informed Neural Networks (PINNs) provide a meshless approach for solving partial differential equations (PDEs), but suffer severe degradation in stiff and shock-dominated problems, where small PDE residuals can correspond to globally inaccurate solutions. We show these failures are multi-causal, arising from the concurrent interplay of (i) spectral bias against sharp features, (ii) imbalanced multi-term optimization and loss-weight...

    arxiv.org/abs/2607.11310 · PDF

  40. 40

    Trustworthy synthetic data for campaign decision support: strategy simulation fidelity and the PolicySynth framework

    Tung Dang, The Hung Phung, Son Lam Nguyen, Tu Nguyen

    cs.LG · stat.ML

    Decision support systems (DSS) increasingly run retention what-if analysis on synthetic customer populations, because privacy constraints preclude unrestricted use of real data. Such a system is trustworthy only if the synthetic data lead managers to the same decisions as the real data would; yet prevailing criteria certify distributional similarity, not decision alignment, so a synthetic population can match every marginal distribution while...

    arxiv.org/abs/2607.11269 · PDF

  41. 41

    FastTPS: An Optimized Method for LLM Token Phase for AI accelerators

    Wenzong Yang, Danyang Zhang, Kun Cao, Tejus Siddagangaiah, Rajeev Patwari, Zhanxing Pu, Siyin Kong, Zijiang Yang,...

    cs.LG

    The popularity of large language models (LLMs) escalates an ongoing demand for effective inference. However, due to the sequential processing of tokens during the token phase in decoder-only LLMs inference, the inherent low parallelism leads to reduced throughput and suboptimal utilization of the computing units on artificial intelligence (AI) accelerators, particularly when handling long-sequence inputs that impose significant memory...

    arxiv.org/abs/2607.11211 · PDF

  42. 42

    NeuroMem-FHP: A Likelihood-Free Deep Learning Framework for Parameter Estimation of Fractional Hawkes Process

    Neha Gupta, Aditya Maheshwari

    cs.LG · stat.ML

    In this paper, we propose deep learning based NeuroMem-FHP framework for estimating the parameters of the fractional Hawkes process (FHP), a self-exciting point process that captures long-range dependence through a fractional Mittag-Leffler excitation kernel. Two neural architectures, namely a Long Short-Term Memory (LSTM) network and a Transformer, are developed to estimate the model parameters $(μ,γ,α,β)$ directly from sequences of...

    arxiv.org/abs/2607.11177 · PDF

  43. 43

    Rank-Conditioned Sample Reuse for the Plackett--Luce Best-of-$K$ Objective

    Melveena Jolly, Midhun Xavier

    cs.LG · stat.ME · stat.ML

    We study the coupled objective J_K^WOR = E_{S ~ PL-WOR_K}[max_{i in S} R_i]: the expected maximum reward of a size-K Plackett-Luce draw without replacement, the law of Gumbel-Top-K / Stochastic Beam Search decoding. This estimand differs from the conventional i.i.d. objective J_K^iid = E[max_{i<=K} R_i] targeted by existing sample-reuse Max@K estimators, and reusing their i.i.d. weights under the coupled sampler is provably biased (a...

    arxiv.org/abs/2607.11146 · PDF

  44. 44

    Learning Subgroup Relations Using Siamese Graph Neural Networks

    Tal Weissblat

    cs.LG · math.GR

    Determining whether one finite group is isomorphic to a subgroup of another is a fundamental problem in computational group theory. In this work, we propose a Siamese Graph Neural Network (Siamese GNN) for subgroup prediction using Cayley graph representations of finite groups. Each input group is represented by its undirected Cayley graph and encoded by one branch of a Siamese GNN to produce a graph embedding. The resulting graph embeddings...

    arxiv.org/abs/2607.11140 · PDF

  45. 45

    ToolAtlas: Learning Once, Reusing Everywhere with Tool-Side Memory

    Yue Fang, Zhibang Yang, Fangkai Yang, Xiaoting Qin, Liqun Li, Qingwei Lin, Saravan Rajmohan, Dongmei Zhang

    cs.LG

    Large language model (LLM) agents increasingly rely on external tools served by shared providers and accessed by heterogeneous downstream agents. Existing approaches improve tool use on the agent side through parameter updates, prompt refinement, or agent-side memory, making tool knowledge difficult to share and limited to behaviors observed in past tasks. We argue that reusable tool knowledge should instead be maintained by the tool...

    arxiv.org/abs/2607.11126 · PDF

  46. 46

    The Equilibrium Is the Initialization: Lazy Identity Collapse in Physics-Structured Deep Equilibrium Reasoning

    Joyjeet Singh

    cs.LG · cs.AI

    Deep equilibrium models promise input-adaptive implicit computation: harder problems should demand more solver iterations, and the solved equilibrium should encode the result of genuine iterative inference. We report a cautionary study of a port-Hamiltonian DEQ with a learned initialization on two reasoning tasks -- ProofWriter entailment over frozen DeBERTa embeddings and a BFS-verified graph-reachability benchmark -- in which the implicit...

    arxiv.org/abs/2607.11116 · PDF

  47. 47

    CA-DGCL: Dynamic Graph Continual Learning via Condensation and Attachment

    Tingxu Yan Ye Yuan

    cs.LG

    Dynamic graph continual learning (DGCL) is an effective manner for handling catastrophic forgetting in dynamic graphs. However, existing DGCL methods underutilize temporal information across graph snapshots. To address this critical issue, we propose a novel framework for Dynamic Graph Continual Learning via Condensation and Attachment (CA-DGCL). Specifically, CA-DGCL first condenses historical graph snapshots into compact semantic...

    arxiv.org/abs/2607.11112 · PDF

  48. 48

    Neural Discovery of Memory and Nonlocal Kernels in Integro-Differential Equations with Constrained Kolmogorov--Arnold Networks

    Aruzhan Tleubek, Salah A Faroughi

    cs.LG · math.NA

    Discovering the memory or nonlocal kernel governing an integro-differential equation (IDE) from sparse and noisy observations is an ill-posed inverse problem. Existing identification methods often rely on problem-specific analytical derivations, specialized observation requirements, or restrictive assumptions about the kernel, limiting their applicability across different classes of IDEs. In this work, we propose a differentiable-solver-based...

    arxiv.org/abs/2607.11110 · PDF

  49. 49

    A Novel Graph Fraud Detector via Grouped Attribute Completion and Confidence-Aware Contrastive Learning

    Junpeng Wu, Ye Yuan

    cs.LG

    Graph fraud detection plays a pivotal role in safeguarding the security and integrity of modern digital ecosystems. Graph Neural Networks (GNNs) are commonly adopted for graph fraud detection. However, the practical performance of existing GNN-based detectors is severely hindered by incomplete node attributes and extreme class imbalance within graphs. To mitigate these limitations, this paper proposes a novel framework for Graph Fraud...

    arxiv.org/abs/2607.11107 · PDF

  50. 50

    Multi-dimensional training-priority weighting based on physical information propagation paths: a unified residual-weighting framework for physics-informed neural networks

    Zhangyi Lian, Xinda Dong, Wenxuan Huo, Weifeng Huang, Gang Zhu, Qiang He

    cs.LG

    Physics-informed neural networks (PINNs) have shown promise for solving partial differential equations (PDEs); however, their synchronous optimization treats residuals of different regions and constraints equally, which is inconsistent with the progressive "from source to response" physical information propagation path, degrading training stability and accuracy. Existing causal training methods focus mainly on the temporal dimension, lacking...

    arxiv.org/abs/2607.11094 · PDF

  51. 51

    Adapting Evidential Neural Networks to Test-Time Neighbor Fusion Improves Molecular Property Prediction

    Cameron Gruich, Weichi Yao, Yixin Wang, Bryan Goldsmith

    cs.LG · q-bio.BM · stat.ML

    A trained molecular property model can be refined at test time by correcting each prediction with the measured labels of the most similar training molecules, a retraining-free procedure we call neighbor fusion; evidential neural networks make it principled by using their aleatoric and epistemic uncertainty to parameterize a Bayesian update. Our main contribution, PG-EVIKAL, learns a property-distance metric to re-rank structurally similar...

    arxiv.org/abs/2607.11091 · PDF

  52. 52

    Link Adaptation Using Joint-Thompson Sampling

    Vignatha Vinjam, Manjunath Kolavennu, Myna Vajha, Karthik Periyapattana Narayanaprasad

    cs.LG

    The choice of Modulation and Coding (MCS) type for a particular channel condition is made through link adaptation (LA) algorithms that operate at the MAC layer. These algorithms rely on the ACK/NACK statistics and the channel quality index (CQI) feedback. Several existing works model LA as a multi-armed bandit (MAB) problem across cellular and Wi-Fi links. In the MAB formulation, each available MCS is a Bernoulli arm parameterized by its...

    arxiv.org/abs/2607.11075 · PDF

  53. 53

    AeroMELD: A Linear Embedding of Aerosol Populations for Diagnostics and Latent Dynamics

    Ehsan Saleh, Saba Ghaffari, Wenhan Tang, Jeffrey H. Curtis, Lekha Patel, Peter A. Bosler, Nicole Riemer, Matthew West

    cs.LG · physics.ao-ph

    Accurately representing atmospheric aerosol populations is essential for simulating aerosol-cloud interactions, radiative forcing, and ice nucleation, yet existing reduced schemes impose structural assumptions that limit their ability to capture composition diversity and mixing state. Machine-learning approaches offer more flexible representations, but standard autoencoders do not preserve the mathematical structure of aerosol populations and...

    arxiv.org/abs/2607.11073 · PDF

  54. 54

    Domain-Aware Scaling Laws Uncover Data Synergy

    Kimia Hamidieh, Lester Mackey, David Alvarez-Melis

    cs.LG · cs.CL

    Machine learning progress is often attributed to scaling model size and dataset volume, yet the composition of data can be just as consequential. Empirical findings repeatedly show that combining datasets from different domains yields nontrivial interactions. For instance, adding code improves mathematical reasoning, while certain mixtures introduce interference that reduces model performance. We refer to these effects collectively as data...

    arxiv.org/abs/2607.11052 · PDF

  55. 55

    When the Reward Suite Is Leaky: A Preregistered Causal Contrast of Natural Verifier False Positives in RLVR

    Chuyifei Zhang

    cs.LG · cs.CL

    The test suites used as RLVR rewards for code have natural false positives: per-task, persistent, asymmetric errors that accept the same wrong programs every time they appear, unlike the symmetric or resampled noise assumed by existing noise-robustness analyses. We run a preregistered two-arm causal contrast on a deployed suite: GRPO on identical MBPP tasks, seeds, and compute, rewarded by the original MBPP tests (leaky) versus the MBPP+...

    arxiv.org/abs/2607.11022 · PDF

  56. 56

    TabPFN beyond Tabular Data: Calibration and Accuracy on Multimodal Embeddings

    Jingxiang Zhang, Lujia Zhong, Zijie Zhu, Shuo Huang, Yuang Xu

    cs.LG

    Few-shot multimodal classification commonly attaches a lightweight head, such as $k$-nearest neighbors, logistic regression, or a linear SVM, to a frozen pretrained encoder. Although computationally efficient, these heads can produce poorly calibrated confidence scores, limiting their reliability in calibration-sensitive applications. We evaluate TabPFN as a plug-and-play, zero-gradient classification head for frozen image, text, and audio...

    arxiv.org/abs/2607.11007 · PDF

  57. 57

    A Multi-Agent Framework for Zero-Dimensional Reduced-Order Model Planning

    Bingteng Sun, Hao Yin, Yiling Chen, Renjie Xiao, Lei Xie, Shanyou Wang, Ruonan Wang, Shubao Chen, Qingzong Xu, Lin...

    cs.LG

    Zero-dimensional reduced-order models (0D ROMs) are central to multi-dimensional design workflows for high-end complex equipment. However, the planning process currently relies on manual expertise, limiting topological exploration and prolonging iterations. Even traditional optimization methods such as Genetic Algorithms (GA) are typically confined to local parameter tuning. Although Large Language Model (LLM) agents have shown promise in...

    arxiv.org/abs/2607.10994 · PDF

  58. 58

    Enhanced Byzantine-Robust Federated Learning Via Truncated-Quadratic Loss for Heterogeneous Data

    Zhi-Yong Wang, Hao Nan Sheng, Werner Stefan, Hing Cheung So, Linqi Song, Weitao Xu

    cs.LG

    Federated learning distributes data among $n$ clients, making it vulnerable to malicious attacks and data heterogeneity, which together pose challenges for robust learning. To tackle this issue, centered clipping and Huber aggregators have been exploited for Byzantine robustness. In this paper, we first demonstrate their equivalence via convex conjugate theory, and show that they can yield biased solutions in the presence of outliers, leading...

    arxiv.org/abs/2607.10970 · PDF

  59. 59

    Efficient Online Proportional Sampling with Applications to Smoothed Online Learning

    Amirmahdi Mirfakhar, Maria-Florina Balcan, Hedyeh Beyhaghi

    cs.LG · cs.AI · cs.CG · cs.GT · econ.TH

    We study the problem of efficient online proportional sampling from a high-dimensional domain under a $σ$-smoothed adversary, where the sampling distribution is induced by a dynamically evolving weight function defined over a sequence of piecewise-structured partitions. This setting captures a broad range of applications, including principal-agent games (e.g., pricing and contract design), and algorithm configuration and parameter tuning. The...

    arxiv.org/abs/2607.10963 · PDF

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.

Colophon Set in Georgia, with system sans for interface chrome and a monospaced stack for code and paper identifiers. Sole accent: amber #D99C5E. Built and served on an always-free VM. The masthead is set 14% letterspaced because newspapers do that and it works.