Cover Image

Uncalibrated Reasoning: GRPO Induces Overconfidence for Stochastic Outcomes

Fuente: arXiv

Saved in:

Bibliographic Details
Main Authors:	Bereket, Michael, Leskovec, Jure
Format:	Preprint
Published:	2025
Subjects:	Machine Learning Artificial Intelligence
Online Access:	Acceder al recurso
Tags:	Add Tag No Tags, Be the first to tag this record!

Similar Items

TimeGraphs: Graph-based Temporal Reasoning
by: Maheshwari, Paridhi, et al.
Published: (2024)

RelGNN: Composite Message Passing for Relational Deep Learning
by: Chen, Tianlang, et al.
Published: (2025)

MLAgentBench: Evaluating Language Agents on Machine Learning Experimentation
by: Huang, Qian, et al.
Published: (2023)

Large Language Models are Good Relational Learners
by: Wu, Fang, et al.
Published: (2025)

Uncertainty Quantification for Forward and Inverse Problems of PDEs via Latent Global Evolution
by: Wu, Tailin, et al.
Published: (2024)

AMIR-GRPO: Inducing Implicit Preference Signals into GRPO
by: Yari, Amir Hossein, et al.
Published: (2026)

Relational Deep Learning: Challenges, Foundations and Next-Generation Architectures
by: Dwivedi, Vijay Prakash, et al.
Published: (2025)

Learning over Positive and Negative Edges with Contrastive Message Passing
by: Pao-Huang, Peter, et al.
Published: (2026)

Modelling Cellular Perturbations with the Sparse Additive Mechanism Shift Variational Autoencoder
by: Bereket, Michael, et al.
Published: (2023)

Agentic Uncertainty Reveals Agentic Overconfidence
by: Kaddour, Jean, et al.
Published: (2026)

GRPO-$λ$: Credit Assignment improves LLM Reasoning
by: Parthasarathi, Prasanna, et al.
Published: (2025)

KumoRFM-2: Scaling Foundation Models for Relational Learning
by: Hudovernik, Valter, et al.
Published: (2026)

S-GRPO: Early Exit via Reinforcement Learning in Reasoning Models
by: Dai, Muzhi, et al.
Published: (2025)

PluRel: Synthetic Data unlocks Scaling Laws for Relational Foundation Models
by: Kothapalli, Vignesh, et al.
Published: (2026)

From Similarity to Superiority: Channel Clustering for Time Series Forecasting
by: Chen, Jialin, et al.
Published: (2024)

Automated Hypothesis Validation with Agentic Sequential Falsifications
by: Huang, Kexin, et al.
Published: (2025)

GRPO is Secretly a Process Reward Model
by: Sullivan, Michael, et al.
Published: (2025)

ExGRPO: Learning to Reason from Experience
by: Zhan, Runzhe, et al.
Published: (2025)

Comparative Analysis and Parametric Tuning of PPO, GRPO, and DAPO for LLM Reasoning Enhancement
by: Lian, Yongsheng
Published: (2025)

Mitigating Overconfidence in Out-of-Distribution Detection by Capturing Extreme Activations
by: Azizmalayeri, Mohammad, et al.
Published: (2024)

Surface-based Molecular Design with Multi-modal Flow Matching
by: Wu, Fang, et al.
Published: (2026)

From Reasoning to Code: GRPO Optimization for Underrepresented Languages
by: Pennino, Federico, et al.
Published: (2025)

TGM: a Modular and Efficient Library for Machine Learning on Temporal Graphs
by: Chmura, Jacob, et al.
Published: (2025)

What is the Alignment Objective of GRPO?
by: Vojnovic, Milan, et al.
Published: (2025)

GRPO-VPS: Enhancing Group Relative Policy Optimization with Verifiable Process Supervision for Effective Reasoning
by: Wang, Jingyi, et al.
Published: (2026)

Multi-Task GRPO: Reliable LLM Reasoning Across Tasks
by: Ramesh, Shyam Sundhar, et al.
Published: (2026)

Overconfident Errors Need Stronger Correction: Asymmetric Confidence Penalties for Reinforcement Learning
by: Xu, Yuanda, et al.
Published: (2026)

Relational Graph Transformer
by: Dwivedi, Vijay Prakash, et al.
Published: (2025)

Predictive Query Language: A Domain-Specific Language for Predictive Modeling on Relational Databases
by: Kocijan, Vid, et al.
Published: (2026)

Stepwise Guided Policy Optimization: Coloring your Incorrect Reasoning in GRPO
by: Chen, Peter, et al.
Published: (2025)

Compositional Generative Inverse Design
by: Wu, Tailin, et al.
Published: (2024)

Relational Transformer: Toward Zero-Shot Foundation Models for Relational Data
by: Ranjan, Rishabh, et al.
Published: (2025)

Scaf-GRPO: Scaffolded Group Relative Policy Optimization for Enhancing LLM Reasoning
by: Zhang, Xichen, et al.
Published: (2025)

GanitLLM: Difficulty-Aware Bengali Mathematical Reasoning through Curriculum-GRPO
by: Dipta, Shubhashis Roy, et al.
Published: (2026)

How Overconfidence in Initial Choices and Underconfidence Under Criticism Modulate Change of Mind in Large Language Models
by: Kumaran, Dharshan, et al.
Published: (2025)

Zero-shot causal learning
by: Nilforoshan, Hamed, et al.
Published: (2023)

Noise-corrected GRPO: From Noisy Rewards to Unbiased Gradients
by: Mansouri, Omar El, et al.
Published: (2025)

A Unified Framework for Rethinking Policy Divergence Measures in GRPO
by: Wu, Qingyuan, et al.
Published: (2026)

Smaller Models are Natural Explorers for Policy-Level Diversity in GRPO
by: Ren, Yiming, et al.
Published: (2026)

Physically Parameterized Differentiable MUSIC for DoA Estimation with Uncalibrated Arrays
by: Chatelier, Baptiste, et al.
Published: (2024)