More Test-Time Compute Can Hurt: Overestimation Bias in LLM Beam Search

Fuente: arXiv

Saved in:

Bibliographic Details
Main Authors:	Dalal, Gal, Hallak, Assaf, Chechik, Gal, Ziser, Yftah
Format:	Preprint
Published:	2026
Subjects:	Machine Learning Artificial Intelligence
Online Access:	Acceder al recurso
Tags:	Add Tag No Tags, Be the first to tag this record!

Similar Items

Policy Gradient with Tree Expansion
by: Dalal, Gal, et al.
Published: (2023)

Policy Optimized Text-to-Image Pipeline Design
by: Gadot, Uri, et al.
Published: (2025)

PlaMo: Plan and Move in Rich 3D Physical Environments
by: Hallak, Assaf, et al.
Published: (2024)

Implementing Reinforcement Learning Datacenter Congestion Control in NVIDIA NICs
by: Fuhrer, Benjamin, et al.
Published: (2022)

AutoLoop: Fast Visual SLAM Fine-tuning through Agentic Curriculum Learning
by: Lahiany, Assaf, et al.
Published: (2025)

From Actions to Words: Towards Abstractive-Textual Policy Summarization in RL
by: Admoni, Sahar, et al.
Published: (2025)

Tree Search-Based Policy Optimization under Stochastic Execution Delay
by: Valensi, David, et al.
Published: (2024)

RL-RC-DoT: A Block-level RL agent for Task-Aware Video Compression
by: Gadot, Uri, et al.
Published: (2025)

Efficient GNN Training Through Structure-Aware Randomized Mini-Batching
by: Balaji, Vignesh, et al.
Published: (2025)

Knowing Before Saying: LLM Representations Encode Information About Chain-of-Thought Success Before Completion
by: Afzal, Anum, et al.
Published: (2025)

Assessing Image Quality Using a Simple Generative Representation
by: Raviv, Simon, et al.
Published: (2024)

Gradient Boosting Reinforcement Learning
by: Fuhrer, Benjamin, et al.
Published: (2024)

Beyond Next Token Probabilities: Learnable, Fast Detection of Hallucinations and Data Contamination on LLM Output Distributions
by: Bar-Shalom, Guy, et al.
Published: (2025)

Add-it: Training-Free Object Insertion in Images With Pretrained Diffusion Models
by: Tewel, Yoad, et al.
Published: (2024)

Spanning the Visual Analogy Space with a Weight Basis of LoRAs
by: Manor, Hila, et al.
Published: (2026)

Training-Free Consistent Text-to-Image Generation
by: Tewel, Yoad, et al.
Published: (2024)

Spectral Editing of Activations for Large Language Model Alignment
by: Qiu, Yifu, et al.
Published: (2024)

Stabilizing Policy Gradients for Sample-Efficient Reinforcement Learning in LLM Reasoning
by: Melo, Luckeciano C., et al.
Published: (2025)

Improved Generalization of Weight Space Networks via Augmentations
by: Shamsian, Aviv, et al.
Published: (2024)

Exploration-Driven Policy Optimization in RLHF: Theoretical Insights on Efficient Data Utilization
by: Du, Yihan, et al.
Published: (2024)

Self-Improving World Modelling with Latent Actions
by: Qiu, Yifu, et al.
Published: (2026)

When More Thinking Hurts: Overthinking in LLM Test-Time Compute Scaling
by: Zhou, Shu, et al.
Published: (2026)

Aligning What LLMs Do and Say: Towards Self-Consistent Explanations
by: Admoni, Sahar, et al.
Published: (2025)

Reliable Policy Iteration: Performance Robustness Across Architecture and Environment Perturbations
by: Eshwar, S. R., et al.
Published: (2025)

WMINet: A Wheel-Mounted Inertial Learning Approach For Mobile-Robot Positioning
by: Versano, Gal, et al.
Published: (2025)

Adversarial Robustness Overestimation and Instability in TRADES
by: Li, Jonathan Weiping, et al.
Published: (2024)

Simple Baselines are Competitive with Code Evolution
by: Gideoni, Yonatan, et al.
Published: (2026)

DDTR: Diffusion Denoising Trace Recovery
by: Matyash, Maximilian, et al.
Published: (2025)

Superscopes: Amplifying Internal Feature Representations for Language Model Interpretation
by: Jacobi, Jonathan, et al.
Published: (2025)

Leveraging Deep Learning for Physical Model Bias of Global Air Quality Estimates
by: Doerksen, Kelsey, et al.
Published: (2025)

Large Language Models for Water Distribution Systems Modeling and Decision-Making
by: Goldshtein, Yinon, et al.
Published: (2025)

Temporal-Difference Variational Continual Learning
by: Melo, Luckeciano C., et al.
Published: (2024)

Split and Conquer Partial Deepfake Speech
by: Rimon, Inbal, et al.
Published: (2026)

Suppressing Overestimation in Q-Learning through Adversarial Behaviors
by: Lee, HyeAnn, et al.
Published: (2023)

Reasoning Introduces New Poisoning Attacks Yet Makes Them More Complicated
by: Foerster, Hanna, et al.
Published: (2025)

Revisiting Meta-Learning with Noisy Labels: Reweighting Dynamics and Theoretical Guarantees
by: Zhang, Yiming, et al.
Published: (2025)

Robust Monocular Visual Odometry using Curriculum Learning
by: Lahiany, Assaf, et al.
Published: (2024)

Key-Locked Rank One Editing for Text-to-Image Personalization
by: Tewel, Yoad, et al.
Published: (2023)

From Glucose Patterns to Health Outcomes: A Generalizable Foundation Model for Continuous Glucose Monitor Data Analysis
by: Lutsker, Guy, et al.
Published: (2024)

The Pitfalls of Memorization: When Memorization Hurts Generalization
by: Bayat, Reza, et al.
Published: (2024)