Image de couverture de livre

Learning Dynamics of VLM Finetuning

Fuente: arXiv

Enregistré dans:

Détails bibliographiques
Auteurs principaux:	Zhang, Jusheng, Cai, Kaitong, Yang, Jing, Wang, Keze
Format:	Preprint
Publié:	2025
Sujets:	Machine Learning Artificial Intelligence
Accès en ligne:	Acceder al recurso
Tags:	Ajouter un tag Pas de tags, Soyez le premier à ajouter un tag!

Documents similaires

Guardian: Decoupling Exploration from Safety in Reinforcement Learning
par: Cai, Kaitong, et autres
Publié: (2025)

CF-VLM:CounterFactual Vision-Language Fine-tuning
par: Zhang, Jusheng, et autres
Publié: (2025)

Causal Invariance and Counterfactual Learning Driven Cooperative Game for Multi-Label Classification
par: Fan, Yijia, et autres
Publié: (2025)

Kolmogorov-Arnold Fourier Networks
par: Zhang, Jusheng, et autres
Publié: (2025)

MAT-Agent: Adaptive Multi-Agent Training Optimization
par: Zhang, Jusheng, et autres
Publié: (2025)

RevFFN: Memory-Efficient Full-Parameter Fine-Tuning of Mixture-of-Experts LLMs with Reversible Blocks
par: Liu, Ningyuan, et autres
Publié: (2025)

Spectral Gating Networks
par: Zhang, Jusheng, et autres
Publié: (2026)

Why Keep Your Doubts to Yourself? Trading Visual Uncertainties in Multi-Agent Bandit Systems
par: Zhang, Jusheng, et autres
Publié: (2026)

Backward-Friendly Optimization: Training Large Language Models with Approximate Gradients under Memory Constraints
par: Yang, Jing, et autres
Publié: (2025)

Rational ANOVA Networks
par: Zhang, Jusheng, et autres
Publié: (2026)

RaCoT: Plug-and-Play Contrastive Example Generation Mechanism for Enhanced LLM Reasoning Reliability
par: Cai, Kaitong, et autres
Publié: (2025)

OSC: Cognitive Orchestration through Dynamic Knowledge Alignment in Multi-Agent LLM Collaboration
par: Zhang, Jusheng, et autres
Publié: (2025)

A Scalable Curiosity-Driven Game-Theoretic Framework for Long-Tail Multi-Label Learning in Data Mining
par: Yang, Jing, et autres
Publié: (2026)

Top-Down Semantic Refinement for Image Captioning
par: Zhang, Jusheng, et autres
Publié: (2025)

HybridToken-VLM: Hybrid Token Compression for Vision-Language Models
par: Zhang, Jusheng, et autres
Publié: (2025)

Cost-Effective Communication: An Auction-based Method for Language Agent Interaction
par: Fan, Yijia, et autres
Publié: (2025)

Failure-Driven Workflow Refinement
par: Zhang, Jusheng, et autres
Publié: (2025)

DrDiff: Dynamic Routing Diffusion with Hierarchical Attention for Breaking the Efficiency-Quality Trade-off
par: Zhang, Jusheng, et autres
Publié: (2025)

Process-of-Thought Reasoning for Videos
par: Zhang, Jusheng, et autres
Publié: (2026)

Agent-GSPO: Communication-Efficient Multi-Agent Systems via Group Sequence Policy Optimization
par: Fan, Yijia, et autres
Publié: (2025)

Learning Dynamics of LLM Finetuning
par: Ren, Yi, et autres
Publié: (2024)

Finetune Once: Decoupling General & Domain Learning with Dynamic Boosted Annealing
par: Tang, Yang, et autres
Publié: (2025)

Representation Finetuning for Continual Learning
par: Luo, Haihua, et autres
Publié: (2026)

Teaching RL Agents to Act Better: VLM as Action Advisor for Online Reinforcement Learning
par: Wu, Xiefeng, et autres
Publié: (2025)

OLLIE: Imitation Learning from Offline Pretraining to Online Finetuning
par: Yue, Sheng, et autres
Publié: (2024)

Dynamics-Predictive Sampling for Active RL Finetuning of Large Reasoning Models
par: Mao, Yixiu, et autres
Publié: (2026)

Holdout-Loss-Based Data Selection for LLM Finetuning via In-Context Learning
par: Zhang, Ling, et autres
Publié: (2025)

Reinforcement Learning Gradients as Vitamin for Online Finetuning Decision Transformers
par: Yan, Kai, et autres
Publié: (2024)

Beyond Parameter Finetuning: Test-Time Representation Refinement for Node Classification
par: Zhang, Jiaxin, et autres
Publié: (2026)

Understanding Hardness of Vision-Language Compositionality from A Token-level Causal Lens
par: Chen, Ziliang, et autres
Publié: (2025)

Mashup Learning: Faster Finetuning by Remixing Past Checkpoints
par: Vaina, Sofia Maria Lo Cicero, et autres
Publié: (2026)

AgriWorld:A World Tools Protocol Framework for Verifiable Agricultural Reasoning with Code-Executing LLM Agents
par: Zhang, Zhixing, et autres
Publié: (2026)

FlashVLM: Text-Guided Visual Token Selection for Large Multimodal Models
par: Cai, Kaitong, et autres
Publié: (2025)

The Impact of Initialization on LoRA Finetuning Dynamics
par: Hayou, Soufiane, et autres
Publié: (2024)

Finetune-Informed Pretraining Boosts Downstream Performance
par: Faysal, Atik, et autres
Publié: (2026)

PTTA: A Pure Text-to-Animation Framework for High-Quality Creation
par: Chen, Ruiqi, et autres
Publié: (2025)

Towards Fast Safe Online Reinforcement Learning via Policy Finetuning
par: Chen, Keru, et autres
Publié: (2024)

On the Emergence of Cross-Task Linearity in the Pretraining-Finetuning Paradigm
par: Zhou, Zhanpeng, et autres
Publié: (2024)

Advancing Autonomous VLM Agents via Variational Subgoal-Conditioned Reinforcement Learning
par: Wu, Qingyuan, et autres
Publié: (2025)

T-REX: Mixture-of-Rank-One-Experts with Semantic-aware Intuition for Multi-task Large Language Model Finetuning
par: Zhang, Rongyu, et autres
Publié: (2024)