Buchumschlag

Thinking Isn't an Illusion: Overcoming the Limitations of Reasoning Models via Tool Augmentations

Fuente: arXiv

Gespeichert in:

Bibliographische Detailangaben
Hauptverfasser:	Song, Zhao, Yue, Song, Zhang, Jiahao
Format:	Preprint
Veröffentlicht:	2025
Schlagworte:	Artificial Intelligence
Online-Zugang:	Acceder al recurso
Tags:	Tag hinzufügen Keine Tags, Fügen Sie den ersten Tag hinzu!

Ähnliche Einträge

When Slower Isn't Truer: Inverse Scaling Law of Truthfulness in Multimodal Reasoning
von: Fang, Sitong, et al.
Veröffentlicht: (2025)

Reasoning Isn't Enough: Examining Truth-Bias and Sycophancy in LLMs
von: Barkett, Emilio, et al.
Veröffentlicht: (2025)

Curveball Steering: The Right Direction To Steer Isn't Always Linear
von: Raval, Shivam, et al.
Veröffentlicht: (2026)

Inverse Scaling: When Bigger Isn't Better
von: McKenzie, Ian R., et al.
Veröffentlicht: (2023)

Comment on The Illusion of Thinking: Understanding the Strengths and Limitations of Reasoning Models via the Lens of Problem Complexity
von: Lawsen, A.
Veröffentlicht: (2025)

Why Isn't Relational Learning Taking Over the World?
von: Poole, David
Veröffentlicht: (2025)

The Illusion of Thinking: Understanding the Strengths and Limitations of Reasoning Models via the Lens of Problem Complexity
von: Shojaee, Parshin, et al.
Veröffentlicht: (2025)

Unlearning Isn't Deletion: Investigating Reversibility of Machine Unlearning in LLMs
von: Xu, Xiaoyu, et al.
Veröffentlicht: (2025)

When Alignment Isn't Enough: Response-Path Attacks on LLM Agents
von: Luo, Mingyu, et al.
Veröffentlicht: (2026)

When Privacy Isn't Synthetic: Hidden Data Leakage in Generative AI Models
von: Mustaqim, S. M., et al.
Veröffentlicht: (2025)

Seeing Isn't Knowing: Do VLMs Know When Not to Answer Spatial Questions (and Why)?
von: Zhang, Yue, et al.
Veröffentlicht: (2026)

Seeing Isn't Believing: Mitigating Belief Inertia via Active Intervention in Embodied Agents
von: Wang, Hanlin, et al.
Veröffentlicht: (2026)

Seeing Isn't Believing: Context-Aware Adversarial Patch Synthesis via Conditional GAN
von: Kazoom, Roie, et al.
Veröffentlicht: (2025)

When Correct Isn't Usable: Improving Structured Output Reliability in Small Language Models
von: Galeone, Cosimo, et al.
Veröffentlicht: (2026)

Rethinking the Illusion of Thinking
von: Varela, Iñaki Dellibarda, et al.
Veröffentlicht: (2025)

When One Moment Isn't Enough: Multi-Moment Retrieval with Cross-Moment Interactions
von: Cao, Zhuo, et al.
Veröffentlicht: (2025)

Talk Isn't Always Cheap: Understanding Failure Modes in Multi-Agent Debate
von: Wynn, Andrea, et al.
Veröffentlicht: (2025)

Recall Isn't Enough: Bounding Commitments in Personalized Language Systems
von: Tang, Rui, et al.
Veröffentlicht: (2026)

Why Synthetic Isn't Real Yet: A Diagnostic Framework for Contact Center Dialogue Generation
von: Devanathan, Rishikesh, et al.
Veröffentlicht: (2025)

More Isn't Always Better: Balancing Decision Accuracy and Conformity Pressures in Multi-AI Advice
von: Tsuchiya, Yuta, et al.
Veröffentlicht: (2026)

Which Coauthor Should I Nominate in My 99 ICLR Submissions? A Mathematical Analysis of the ICLR 2026 Reciprocal Reviewer Nomination Policy
von: Song, Zhao, et al.
Veröffentlicht: (2025)

Think How to Think: Mitigating Overthinking with Autonomous Difficulty Cognition in Large Reasoning Models
von: Liu, Yongjiang, et al.
Veröffentlicht: (2025)

Evaluating Frontier LLMs on PhD-Level Mathematical Reasoning: A Benchmark on a Textbook in Theoretical Computer Science about Randomized Algorithms
von: Cao, Yang, et al.
Veröffentlicht: (2025)

When Pretty Isn't Useful: Investigating Why Modern Text-to-Image Models Fail as Reliable Training Data Generators
von: Adamkiewicz, Krzysztof, et al.
Veröffentlicht: (2026)

The Illusion of Insight in Reasoning Models
von: d'Aliberti, Liv G., et al.
Veröffentlicht: (2026)

A Comment On "The Illusion of Thinking": Reframing the Reasoning Cliff as an Agentic Gap
von: Khan, Sheraz, et al.
Veröffentlicht: (2025)

An Empirical Study of Reasoning Steps in Thinking Code LLMs
von: Xue, Haoran, et al.
Veröffentlicht: (2025)

The Tool-Overuse Illusion: Why Does LLM Prefer External Tools over Internal Knowledge?
von: Zeng, Yirong, et al.
Veröffentlicht: (2026)

AdapThink: Adaptive Thinking Preferences for Reasoning Language Model
von: Wan, Xu, et al.
Veröffentlicht: (2025)

Breaking the Illusion of Identity in LLM Tooling
von: Miller, Marek
Veröffentlicht: (2026)

Thinking with Comics: Enhancing Multimodal Reasoning through Structured Visual Storytelling
von: Chen, Andong, et al.
Veröffentlicht: (2026)

ThinkPilot: Steering Reasoning Models via Automated Think-prefixes Optimization
von: Li, Sunzhu, et al.
Veröffentlicht: (2025)

Thinking by Doing: Building Efficient World Model Reasoning in LLMs via Multi-turn Interaction
von: Shu, Bao, et al.
Veröffentlicht: (2025)

KAG-Thinker: Interactive Thinking and Deep Reasoning in LLMs via Knowledge-Augmented Generation
von: Zhang, Dalong, et al.
Veröffentlicht: (2025)

OPE: Overcoming Information Saturation in Parallel Thinking via Outline-Guided Path Exploration
von: Guo, Qi, et al.
Veröffentlicht: (2026)

Re-Tuning: Overcoming the Compositionality Limits of Large Language Models with Recursive Tuning
von: Pasewark, Eric, et al.
Veröffentlicht: (2024)

ToolMind Technical Report: A Large-Scale, Reasoning-Enhanced Tool-Use Dataset
von: Yang, Chen, et al.
Veröffentlicht: (2025)

Multi-Step Reasoning for Embodied Question Answering via Tool Augmentation
von: Zhai, Mingliang, et al.
Veröffentlicht: (2025)

Provable Failure of Language Models in Learning Majority Boolean Logic via Gradient Descent
von: Chen, Bo, et al.
Veröffentlicht: (2025)

Knowing Isn't Understanding: Re-grounding Generative Proactivity with Epistemic and Behavioral Insight
von: Kaur, Kirandeep, et al.
Veröffentlicht: (2026)