Image de couverture de livre

The Self-Improvement Paradox: Can Language Models Bootstrap Reasoning Capabilities without External Scaffolding?

Fuente: arXiv

Enregistré dans:

Détails bibliographiques
Auteurs principaux:	Sun, Yutao, Chen, Mingshuai, Zhao, Tiancheng, Xu, Ruochen, Zhang, Zilun, Yin, Jianwei
Format:	Preprint
Publié:	2025
Sujets:	Computation and Language Artificial Intelligence
Accès en ligne:	Acceder al recurso
Tags:	Ajouter un tag Pas de tags, Soyez le premier à ajouter un tag!

Documents similaires

Preserving Knowledge in Large Language Model with Model-Agnostic Self-Decompression
par: Zhang, Zilun, et autres
Publié: (2024)

Talking to Yourself: Defying Forgetting in Large Language Models
par: Sun, Yutao, et autres
Publié: (2026)

RS5M and GeoRSCLIP: A Large Scale Vision-Language Dataset and A Large Vision-Language Model for Remote Sensing
par: Zhang, Zilun, et autres
Publié: (2023)

HORAE: A Domain-Agnostic Language for Automated Service Regulation
par: Sun, Yutao, et autres
Publié: (2024)

ZoomEye: Enhancing Multimodal LLMs with Human-Like Zooming Capabilities through Tree-Based Image Exploration
par: Shen, Haozhan, et autres
Publié: (2024)

Thought Manipulation: External Thought Can Be Efficient for Large Reasoning Models
par: Liu, Yule, et autres
Publié: (2025)

TextGames: Learning to Self-Play Text-Based Puzzle Games via Language Model Reasoning
par: Hudi, Frederikus, et autres
Publié: (2025)

RCP-Merging: Merging Long Chain-of-Thought Models with Domain-Specific Models by Considering Reasoning Capability as Prior
par: Yang, Junyao, et autres
Publié: (2025)

BRiTE: Bootstrapping Reinforced Thinking Process to Enhance Language Model Reasoning
par: Zhong, Han, et autres
Publié: (2025)

SciAgent: Tool-augmented Language Models for Scientific Reasoning
par: Ma, Yubo, et autres
Publié: (2024)

Large Language Models Can Self-Improve in Long-context Reasoning
par: Li, Siheng, et autres
Publié: (2024)

GeoRSMLLM: A Multimodal Large Language Model for Vision-Language Tasks in Geoscience and Remote Sensing
par: Zhang, Zilun, et autres
Publié: (2025)

LightReasoner: Can Small Language Models Teach Large Language Models Reasoning?
par: Wang, Jingyuan, et autres
Publié: (2025)

MeTHanol: Modularized Thinking Language Models with Intermediate Layer Thinking, Decoding and Bootstrapping Reasoning
par: Xi, Ningyuan, et autres
Publié: (2024)

OpenToM: A Comprehensive Benchmark for Evaluating Theory-of-Mind Reasoning Capabilities of Large Language Models
par: Xu, Hainiu, et autres
Publié: (2024)

Test-time Recursive Thinking: Self-Improvement without External Feedback
par: Zhuang, Yufan, et autres
Publié: (2026)

Language Models are Hidden Reasoners: Unlocking Latent Reasoning Capabilities via Self-Rewarding
par: Chen, Haolin, et autres
Publié: (2024)

Can They Dixit? Yes they Can! Dixit as a Playground for Multimodal Language Model Capabilities
par: Balepur, Nishant, et autres
Publié: (2025)

TTSR: Test-Time Self-Reflection for Continual Reasoning Improvement
par: He, Haoyang, et autres
Publié: (2026)

WebThinker: Empowering Large Reasoning Models with Deep Research Capability
par: Li, Xiaoxi, et autres
Publié: (2025)

Reasoning Capabilities of Large Language Models on Dynamic Tasks
par: Wong, Annie, et autres
Publié: (2025)

Evaluating Consistency and Reasoning Capabilities of Large Language Models
par: Saxena, Yash, et autres
Publié: (2024)

Distilling Mathematical Reasoning Capabilities into Small Language Models
par: Zhu, Xunyu, et autres
Publié: (2024)

Enhancing Chain-of-Thoughts Prompting with Iterative Bootstrapping in Large Language Models
par: Sun, Jiashuo, et autres
Publié: (2023)

Deceptive Semantic Shortcuts on Reasoning Chains: How Far Can Models Go without Hallucination?
par: Li, Bangzheng, et autres
Publié: (2023)

Can Large Language Models Understand, Reason About, and Generate Code-Switched Text?
par: Winata, Genta Indra, et autres
Publié: (2026)

The Accuracy Paradox in RLHF: When Better Reward Models Don't Yield Better Language Models
par: Chen, Yanjun, et autres
Publié: (2024)

Can't say cant? Measuring and Reasoning of Dark Jargons in Large Language Models
par: Ji, Xu, et autres
Publié: (2024)

SafeChain: Safety of Language Models with Long Chain-of-Thought Reasoning Capabilities
par: Jiang, Fengqing, et autres
Publié: (2025)

Confidence Matters: Revisiting Intrinsic Self-Correction Capabilities of Large Language Models
par: Li, Loka, et autres
Publié: (2024)

Unlocking Reasoning Capability on Machine Translation in Large Language Models
par: Rajaee, Sara, et autres
Publié: (2026)

CHANCERY: Evaluating Corporate Governance Reasoning Capabilities in Language Models
par: Irwin, Lucas, et autres
Publié: (2025)

Self-Improvement in Language Models: The Sharpening Mechanism
par: Huang, Audrey, et autres
Publié: (2024)

Vision-Language Models Can Self-Improve Reasoning via Reflection
par: Cheng, Kanzhi, et autres
Publié: (2024)

Reasoning or Reciting? Exploring the Capabilities and Limitations of Language Models Through Counterfactual Tasks
par: Wu, Zhaofeng, et autres
Publié: (2023)

TReB: A Comprehensive Benchmark for Evaluating Table Reasoning Capabilities of Large Language Models
par: Li, Ce, et autres
Publié: (2025)

LLMs are Superior Feedback Providers: Bootstrapping Reasoning for Lie Detection with Self-Generated Feedback
par: Banerjee, Tanushree, et autres
Publié: (2024)

Solving for X and Beyond: Can Large Language Models Solve Complex Math Problems with More-Than-Two Unknowns?
par: Kao, Kuei-Chun, et autres
Publié: (2024)

SEKI: Self-Evolution and Knowledge Inspiration based Neural Architecture Search via Large Language Models
par: Cai, Zicheng, et autres
Publié: (2025)

Self-Route: Automatic Mode Switching via Capability Estimation for Efficient Reasoning
par: He, Yang, et autres
Publié: (2025)