Tootfinder

HardcoreLogic: Challenging Large Reasoning Models with Long-tail Logic Puzzle Games
Jingcong Liang, Shijun Wan, Xuehai Wu, Siyuan Wang, Yitong Li, Qianglong Chen, Duyu Tang, Zhongyu Wei
https://arxiv.org/abs/2510.12563

HardcoreLogic: Challenging Large Reasoning Models with Long-tail Logic Puzzle Games
Large Reasoning Models (LRMs) have demonstrated impressive performance on complex tasks, including logical puzzle games that require deriving solutions satisfying all constraints. However, whether they can flexibly apply appropriate rules to varying conditions, particularly when faced with non-canonical game variants, remains an open question. Existing corpora focus on popular puzzles like 9x9 Sudoku, risking overfitting to canonical formats and memorization of solution patterns, which can mask…

Attack via Overfitting: 10-shot Benign Fine-tuning to Jailbreak LLMs
Zhixin Xie, Xurui Song, Jun Luo
https://arxiv.org/abs/2510.02833 https://arxiv.org/pdf…

Attack via Overfitting: 10-shot Benign Fine-tuning to Jailbreak LLMs
Despite substantial efforts in safety alignment, recent research indicates that Large Language Models (LLMs) remain highly susceptible to jailbreak attacks. Among these attacks, finetuning-based ones that compromise LLMs' safety alignment via fine-tuning stand out due to its stable jailbreak performance. In particular, a recent study indicates that fine-tuning with as few as 10 harmful question-answer (QA) pairs can lead to successful jailbreaking across various harmful questions. However, such…

A look at various quirks in AI-generated prose, mainly influenced by "overfitting" in AI models, as humans increasingly mimic AI language in writing and speech (Sam Kriss/New York Times)
https://www.ny…

Why Does A.I. Write Like … That?
If only they were robotic! Instead, chatbots have developed a distinctive — and grating — voice.

Risk Phase Transitions in Spiked Regression: Alignment Driven Benign and Catastrophic Overfitting
Jiping Li, Rishi Sonthalia
https://arxiv.org/abs/2510.01414 https://

Risk Phase Transitions in Spiked Regression: Alignment Driven Benign and Catastrophic Overfitting
This paper analyzes the generalization error of minimum-norm interpolating solutions in linear regression using spiked covariance data models. The paper characterizes how varying spike strengths and target-spike alignments can affect risk, especially in overparameterized settings. The study presents an exact expression for the generalization error, leading to a comprehensive classification of benign, tempered, and catastrophic overfitting regimes based on spike strength, the aspect ratio $c=d/n…

Crosslisted article(s) found for cs.CV. https://arxiv.org/list/cs.CV/new
[1/2]:
- DUA-D2C: Dynamic Uncertainty Aware Method for Overfitting Remediation in Deep Learning
Md. Saiful Bari Siddiqui, Md Mohaiminul Islam, Md. Golam Rabiul Alam

MatheMagic: Generating Dynamic Mathematics Benchmarks Robust to Memorization
Dayy\'an O'Brien, Barry Haddow, Emily Allaway, Pinzhen Chen
https://arxiv.org/abs/2510.05962

MatheMagic: Generating Dynamic Mathematics Benchmarks Robust to Memorization
Conducting contamination-free evaluation of mathematical capabilities can be difficult for two reasons: models may memorize a test set once it is made public, and current mathematical benchmarks are prone to overfitting due to having limited diversity of symbols and rules, coupled with closed-ended answers. This paper proposes a method to leverage these shortcomings as useful features to a construct dynamic, counterfactual benchmark, which can be used to both reveal overfitting and measure true…

Crosslisted article(s) found for cs.NE. https://arxiv.org/list/cs.NE/new
[1/1]:
- DUA-D2C: Dynamic Uncertainty Aware Method for Overfitting Remediation in Deep Learning
Md. Saiful Bari Siddiqui, Md Mohaiminul Islam, Md. Golam Rabiul Alam

Overfitting in Adaptive Robust Optimization
Karl Zhu, Dimitris Bertsimas
https://arxiv.org/abs/2509.16451 https://arxiv.org/pdf/2509.16451

Overfitting in Adaptive Robust Optimization
Adaptive robust optimization (ARO) extends static robust optimization by allowing decisions to depend on the realized uncertainty - weakly dominating static solutions within the modeled uncertainty set. However, ARO makes previous constraints that were independent of uncertainty now dependent, making it vulnerable to additional infeasibilities when realizations fall outside the uncertainty set. This phenomenon of adaptive policies being brittle is analogous to overfitting in machine learning. T…

Scarce Data, Noisy Inferences, and Overfitting: The Hidden Flaws in Ecological Dynamics Modelling
Mario Castro, Rafael Vida, Javier Galeano, Jos\'e A. Cuesta
https://arxiv.org/abs/2510.03718

Scarce Data, Noisy Inferences, and Overfitting: The Hidden Flaws in Ecological Dynamics Modelling
Metagenomic data has significantly advanced microbiome research by employing ecological models, particularly in personalised medicine. The generalised Lotka-Volterra (gLV) model is commonly used to understand microbial interactions and predict ecosystem dynamics. However, gLV models often fail to capture complex interactions, especially when data is limited or noisy. This study critically assesses the effectiveness of gLV and similar models using Bayesian inference and a model reduction method …

Hyperspectral data augmentation with transformer-based diffusion models
Mattia Ferrari, Lorenzo Bruzzone
https://arxiv.org/abs/2510.08363 https://arxiv.org…

Hyperspectral data augmentation with transformer-based diffusion models
The introduction of new generation hyperspectral satellite sensors, combined with advancements in deep learning methodologies, has significantly enhanced the ability to discriminate detailed land-cover classes at medium-large scales. However, a significant challenge in deep learning methods is the risk of overfitting when training networks with small labeled datasets. In this work, we propose a data augmentation technique that leverages a guided diffusion model. To effectively train the model w…

Crosslisted article(s) found for stat.ML. https://arxiv.org/list/stat.ML/new
[1/1]:
- DUA-D2C: Dynamic Uncertainty Aware Method for Overfitting Remediation in Deep Learning
Md. Saiful Bari Siddiqui, Md Mohaiminul Islam, Md. Golam Rabiul Alam

Crosslisted article(s) found for cs.AI. https://arxiv.org/list/cs.AI/new
[1/8]:
- DUA-D2C: Dynamic Uncertainty Aware Method for Overfitting Remediation in Deep Learning
Md. Saiful Bari Siddiqui, Md Mohaiminul Islam, Md. Golam Rabiul Alam

D$^2$GS: Depth-and-Density Guided Gaussian Splatting for Stable and Accurate Sparse-View Reconstruction
Meixi Song, Xin Lin, Dizhe Zhang, Haodong Li, Xiangtai Li, Bo Du, Lu Qi
https://arxiv.org/abs/2510.08566

D$^2$GS: Depth-and-Density Guided Gaussian Splatting for Stable and Accurate Sparse-View Reconstruction
Recent advances in 3D Gaussian Splatting (3DGS) enable real-time, high-fidelity novel view synthesis (NVS) with explicit 3D representations. However, performance degradation and instability remain significant under sparse-view conditions. In this work, we identify two key failure modes under sparse-view conditions: overfitting in regions with excessive Gaussian density near the camera, and underfitting in distant areas with insufficient Gaussian coverage. To address these challenges, we propose…

Crosslisted article(s) found for physics.bio-ph. https://arxiv.org/list/physics.bio-ph/new
[1/1]:
- Scarce Data, Noisy Inferences, and Overfitting: The Hidden Flaws in Ecological Dynamics Modelling
Mario Castro, Rafael Vida, Javier Galeano, Jos\'e A. Cuesta

R-Log: Incentivizing Log Analysis Capability in LLMs via Reasoning-based Reinforcement Learning
Yilun Liu, Ziang Chen, Song Xu, Minggui He, Shimin Tao, Weibin Meng, Yuming Xie, Tao Han, Chunguang Zhao, Jingzhou Du, Daimeng Wei, Shenglin Zhang, Yongqian Sun
https://arxiv.org/abs/2509.25987

R-Log: Incentivizing Log Analysis Capability in LLMs via Reasoning-based Reinforcement Learning
The growing complexity of log data in modern software systems has prompted the use of Large Language Models (LLMs) for automated log analysis. Current approaches typically rely on direct supervised fine-tuning (SFT) on log-label pairs. However, this exacerbates the domain discrepancy between general-purpose LLMs and specialized log data, causing overfitting. Furthermore, SFT's imbalanced loss computation often allows lengthy contexts to overwhelm critical, concise details in model answers, lead…

Poison to Detect: Detection of Targeted Overfitting in Federated Learning
Soumia Zohra El Mestari, Maciej Krzysztof Zuziak, Gabriele Lenzini
https://arxiv.org/abs/2509.11974 htt…

Poison to Detect: Detection of Targeted Overfitting in Federated Learning
Federated Learning (FL) enables collaborative model training across decentralised clients while keeping local data private, making it a widely adopted privacy-enhancing technology (PET). Despite its privacy benefits, FL remains vulnerable to privacy attacks, including those targeting specific clients. In this paper, we study an underexplored threat where a dishonest orchestrator intentionally manipulates the aggregation process to induce targeted overfitting in the local models of specific clie…

Crosslisted article(s) found for cs.LG. https://arxiv.org/list/cs.LG/new
[2/7]:
- Overfitting in Adaptive Robust Optimization
Karl Zhu, Dimitris Bertsimas
https:/…

Do You Need Proprioceptive States in Visuomotor Policies?
Juntu Zhao, Wenbo Lu, Di Zhang, Yufeng Liu, Yushen Liang, Tianluo Zhang, Yifeng Cao, Junyuan Xie, Yingdong Hu, Shengjie Wang, Junliang Guo, Dequan Wang, Yang Gao
https://arxiv.org/abs/2509.18644

Minimum Sample Size Calculation for Multivariable Regression of Continuous Outcomes in Chemometrics for Astrobiology and Planetary Science
M. Konstantinidis, E. A. Lalla, S. J. Gonzalez, J. Manrique, G. Lopez-Reyes, A. Barlow, E. Sawyers, B. Barrios, M. G. Daly
https://arxiv.org/abs/2510.01210

Minimum Sample Size Calculation for Multivariable Regression of Continuous Outcomes in Chemometrics for Astrobiology and Planetary Science
Over the last few decades, prediction models have become a fundamental tool in statistics, chemometrics, and related fields. However, to ensure that such models have high value, the inferences that they generate must be reliable. In this regard, the internal validity of a prediction model might be threatened if it is not calibrated with a sufficiently large sample size, as problems such as overfitting may occur. Such situations would be highly problematic in many fields, including space science…

Reward Model Routing in Alignment
Xinle Wu, Yao Lu
https://arxiv.org/abs/2510.02850 https://arxiv.org/pdf/2510.02850

Reward Model Routing in Alignment
Reinforcement learning from human or AI feedback (RLHF / RLAIF) has become the standard paradigm for aligning large language models (LLMs). However, most pipelines rely on a single reward model (RM), limiting alignment quality and risking overfitting. Recent work explores RM routing--dynamically selecting an RM from a candidate pool to exploit complementary strengths while maintaining $O(1)$ RM calls--but existing methods suffer from cold-start and insufficient exploration. We propose BayesianR…

Sample Size Calculations for the Development of Risk Prediction Models that Account for Performance Variability
Menelaos Pavlou, Rumana Z. Omar, Gareth Ambler
https://arxiv.org/abs/2509.14028

Sample Size Calculations for the Development of Risk Prediction Models that Account for Performance Variability
Existing approaches to sample size calculations for developing clinical prediction models have focused on ensuring that the expected value of a chosen performance measure meets a pre-specified target. For example, to limit model-overfitting, the sample size is commonly chosen such that the expected calibration slope (CS) is 0.9, close to 1 for a perfectly calibrated model. In practice, due to sampling variability, model performance can vary considerably across different development samples of t…

AudioFuse: Unified Spectral-Temporal Learning via a Hybrid ViT-1D CNN Architecture for Robust Phonocardiogram Classification
Md. Saiful Bari Siddiqui, Utsab Saha
https://arxiv.org/abs/2509.23454

AudioFuse: Unified Spectral-Temporal Learning via a Hybrid ViT-1D CNN Architecture for Robust Phonocardiogram Classification
Biomedical audio signals, such as phonocardiograms (PCG), are inherently rhythmic and contain diagnostic information in both their spectral (tonal) and temporal domains. Standard 2D spectrograms provide rich spectral features but compromise the phase information and temporal precision of the 1D waveform. We propose AudioFuse, an architecture that simultaneously learns from both complementary representations to classify PCGs. To mitigate the overfitting risk common in fusion models, we integrate…

AD-GS: Alternating Densification for Sparse-Input 3D Gaussian Splatting
Gurutva Patle, Nilay Girgaonkar, Nagabhushan Somraj, Rajiv Soundararajan
https://arxiv.org/abs/2509.11003

AD-GS: Alternating Densification for Sparse-Input 3D Gaussian Splatting
3D Gaussian Splatting (3DGS) has shown impressive results in real-time novel view synthesis. However, it often struggles under sparse-view settings, producing undesirable artifacts such as floaters, inaccurate geometry, and overfitting due to limited observations. We find that a key contributing factor is uncontrolled densification, where adding Gaussian primitives rapidly without guidance can harm geometry and cause artifacts. We propose AD-GS, a novel alternating densification framework that …

CollabVLA: Self-Reflective Vision-Language-Action Model Dreaming Together with Human
Nan Sun, Yongchang Li, Chenxu Wang, Huiying Li, Huaping Liu
https://arxiv.org/abs/2509.14889

CollabVLA: Self-Reflective Vision-Language-Action Model Dreaming Together with Human
In this work, we present CollabVLA, a self-reflective vision-language-action framework that transforms a standard visuomotor policy into a collaborative assistant. CollabVLA tackles key limitations of prior VLAs, including domain overfitting, non-interpretable reasoning, and the high latency of auxiliary generative models, by integrating VLM-based reflective reasoning with diffusion-based action generation under a mixture-of-experts design. Through a two-stage training recipe of action groundin…

Convex Regression with a Penalty
Eunji Lim
https://arxiv.org/abs/2509.19788 https://arxiv.org/pdf/2509.19788

Convex Regression with a Penalty
A common way to estimate an unknown convex regression function $f_0: Ω\subset \mathbb{R}^d \rightarrow \mathbb{R}$ from a set of $n$ noisy observations is to fit a convex function that minimizes the sum of squared errors. However, this estimator is known for its tendency to overfit near the boundary of $Ω$, posing significant challenges in real-world applications. In this paper, we introduce a new estimator of $f_0$ that avoids this overfitting by minimizing a penalty on the subgradient while…

MoTiC: Momentum Tightness and Contrast for Few-Shot Class-Incremental Learning
Zeyu He, Shuai Huang, Yuwu Lu, Ming Zhao
https://arxiv.org/abs/2509.19664 https://

MoTiC: Momentum Tightness and Contrast for Few-Shot Class-Incremental Learning
Few-Shot Class-Incremental Learning (FSCIL) must contend with the dual challenge of learning new classes from scarce samples while preserving old class knowledge. Existing methods use the frozen feature extractor and class-averaged prototypes to mitigate against catastrophic forgetting and overfitting. However, new-class prototypes suffer significant estimation bias due to extreme data scarcity, whereas base-class prototypes benefit from sufficient data. In this work, we theoretically demonstra…

RAPTOR: A Foundation Policy for Quadrotor Control
Jonas Eschmann, Dario Albani, Giuseppe Loianno
https://arxiv.org/abs/2509.11481 https://arxiv.org/pdf/250…

RAPTOR: A Foundation Policy for Quadrotor Control
Humans are remarkably data-efficient when adapting to new unseen conditions, like driving a new car. In contrast, modern robotic control systems, like neural network policies trained using Reinforcement Learning (RL), are highly specialized for single environments. Because of this overfitting, they are known to break down even under small differences like the Simulation-to-Reality (Sim2Real) gap and require system identification and retraining for even minimal changes to the system. In this wor…