$Cover Image$

Understanding and Mitigating Spurious Signal Amplification in Test-Time Reinforcement Learning for Math Reasoning

Fuente: arXiv

Saved in:

Bibliographic Details
Main Authors:	Yu, Yongcan, He, Lingxiao, Liang, Jian, Guo, Kuangpu, Wang, Meng, Xie, Qianlong, Wang, Xingxing, He, Ran
Format:	Preprint
Published:	2026
Subjects:	Machine Learning Artificial Intelligence Computation and Language
Online Access:	Acceder al recurso
Tags:	Add Tag No Tags, Be the first to tag this record!

Similar Items

Reassessing the Role of Supervised Fine-Tuning: An Empirical Study in VLM Reasoning
by: Yu, Yongcan, et al.
Published: (2025)

Cooperative Pseudo Labeling for Unsupervised Federated Classification
by: Guo, Kuangpu, et al.
Published: (2025)

Test-Time Immunization: A Universal Defense Framework Against Jailbreaks for (Multimodal) Large Language Models
by: Yu, Yongcan, et al.
Published: (2025)

STAMP: Outlier-Aware Test-Time Adaptation with Stable Memory Replay
by: Yu, Yongcan, et al.
Published: (2024)

Do MLLMs Really Understand Space? A Mathematical Reasoning Evaluation
by: Lu, Shuo, et al.
Published: (2026)

A Comprehensive Survey on Trustworthiness in Reasoning with Large Language Models
by: Wang, Yanbo, et al.
Published: (2025)

Mitigating the Safety-utility Trade-off in LLM Alignment via Adaptive Safe Context Learning
by: Wang, Yanbo, et al.
Published: (2026)

Stay Unique, Stay Efficient: Preserving Model Personality in Multi-Task Merging
by: Guo, Kuangpu, et al.
Published: (2025)

Exploring Vacant Classes in Label-Skewed Federated Learning
by: Guo, Kuangpu, et al.
Published: (2024)

How to Train Your Deep Research Agent? Prompt, Reward, and Policy Optimization in Search-R1
by: Xu, Yinuo, et al.
Published: (2026)

Safe Offline Reinforcement Learning with Real-Time Budget Constraints
by: Lin, Qian, et al.
Published: (2023)

What If Consensus Lies? Selective-Complementary Reinforcement Learning at Test Time
by: Yan, Dong, et al.
Published: (2026)

An Enhanced Federated Prototype Learning Method under Domain Shift
by: Kuang, Liang, et al.
Published: (2024)

Amplification Effects in Test-Time Reinforcement Learning: Safety and Reasoning Vulnerabilities
by: Khattar, Vanshaj, et al.
Published: (2026)

Personalized Federated Learning via Dual-Prompt Optimization and Cross Fusion
by: Zhang, Yuguang, et al.
Published: (2025)

Off-Policy Primal-Dual Safe Reinforcement Learning
by: Wu, Zifan, et al.
Published: (2024)

A Comprehensive Survey on Test-Time Adaptation under Distribution Shifts
by: Liang, Jian, et al.
Published: (2023)

TabularMath: Understanding Math Reasoning over Tables with Large Language Models
by: Tian, Shi-Yu, et al.
Published: (2025)

R-TPT: Improving Adversarial Robustness of Vision-Language Models through Test-Time Prompt Tuning
by: Sheng, Lijun, et al.
Published: (2025)

CodeReasoner: Enhancing the Code Reasoning Ability with Reinforcement Learning
by: Tang, Lingxiao, et al.
Published: (2025)

MathScale: Scaling Instruction Tuning for Mathematical Reasoning
by: Tang, Zhengyang, et al.
Published: (2024)

RL-MPCA: A Reinforcement Learning Based Multi-Phase Computation Allocation Approach for Recommender Systems
by: Zhou, Jiahong, et al.
Published: (2023)

AUTO: Adaptive Outlier Optimization for Test-Time OOD Detection
by: Yang, Puning, et al.
Published: (2023)

Reasoning in Transformers -- Mitigating Spurious Correlations and Reasoning Shortcuts
by: Enström, Daniel, et al.
Published: (2024)

The Illusion of Progress? A Critical Look at Test-Time Adaptation for Vision-Language Models
by: Sheng, Lijun, et al.
Published: (2025)

FITRep: Attention-Guided Item Representation via MLLMs
by: Zhang, Guoxiao, et al.
Published: (2025)

NFT: Bridging Supervised Learning and Reinforcement Learning in Math Reasoning
by: Chen, Huayu, et al.
Published: (2025)

The Trap of Trajectory: Towards Understanding and Mitigating Spurious Correlations in Agentic Memory
by: Tang, Luoxi, et al.
Published: (2026)

Mitigating Spurious Correlations with Causal Logit Perturbation
by: Zhou, Xiaoling, et al.
Published: (2025)

Understanding and Mitigating Spurious Correlations in Text Classification with Neighborhood Analysis
by: Chew, Oscar, et al.
Published: (2023)

HiBid: A Cross-Channel Constrained Bidding System with Budget Allocation by Hierarchical Offline Deep Reinforcement Learning
by: Wang, Hao, et al.
Published: (2023)

DeepResearch-Slice: Bridging the Retrieval-Utilization Gap via Explicit Text Slicing
by: Lu, Shuo, et al.
Published: (2025)

STAPO: Stabilizing Reinforcement Learning for LLMs by Silencing Rare Spurious Tokens
by: Liu, Shiqi, et al.
Published: (2026)

Generative Large-Scale Pre-trained Models for Automated Ad Bidding Optimization
by: Lei, Yu, et al.
Published: (2025)

Mitigating Spurious Correlations for Self-supervised Recommendation
by: Lin, Xinyu, et al.
Published: (2022)

Precedent-Informed Reasoning: Mitigating Overthinking in Large Reasoning Models via Test-Time Precedent Learning
by: Wang, Qianyue, et al.
Published: (2026)

Towards Eliminating Hard Label Constraints in Gradient Inversion Attacks
by: Wang, Yanbo, et al.
Published: (2024)

VideoChat-R1.5: Visual Test-Time Scaling to Reinforce Multimodal Reasoning by Iterative Perception
by: Yan, Ziang, et al.
Published: (2025)

Spurious Correlations and Beyond: Understanding and Mitigating Shortcut Learning in SDOH Extraction with Large Language Models
by: Sakib, Fardin Ahsan, et al.
Published: (2025)

Elastic Representation: Mitigating Spurious Correlations for Group Robustness
by: Wen, Tao, et al.
Published: (2025)