Aligning Large Vision-Language Models at Test Time: A Trajectory-Guided Structured Sampling Approach
Tianbao Jiang, Weicong Ni, Gerard de Melo, Linlin Wang
cs.CL, cs.AI
Submitted: 2026-08-04
License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Terminology
Sources
- Phi-4 Technical Report
- Qwen3-VL Technical Report
- Qwen2.5-VL Technical Report
- Upper and lower bounds on the subgeometric convergence of adaptive Markov chain Monte Carlo
- Expanding Performance Boundaries of Open-Source Multimodal Models with Model, Data, and Test-Time Scaling
- The Entropy Mechanism of Reinforcement Learning for Reasoning Language Models
- OpenVLThinker: Complex Vision-Language Reasoning via Iterative SFT-RL Cycles
- Vision-R1: Incentivizing Reasoning Capability in Multimodal Large Language Models
- Test-time reward-guided alignment of language models by importance sampling on pre-logit space
- Reasoning with Sampling: Your Base Model is Smarter Than You Think
- ThinkPilot: Steering Reasoning Models via Automated Think-prefixes Optimization
- Self-Rewarding Vision-Language Model via Reasoning Decomposition
- OctoTools: An Agentic Framework with Extensible Tools for Complex Reasoning
- MM-Eureka: Exploring the Frontiers of Multimodal Reasoning with Rule-based Reinforcement Learning
- More Thought, Less Accuracy? On the Dual Nature of Reasoning in Vision-Language Models
- Test-Time Alignment for Large Language Models via Textual Model Predictive Control
- SoTA with Less: MCTS-Guided Sample Selection for Data-Efficient Visual Reasoning Self-Improvement
- AStar: Boosting Multimodal Reasoning with Automated Structured Thinking
- GenARM: Reward Guided Generation with Autoregressive Reward Model for Test-time Alignment
- MathVerse: Does Your Multi-modal LLM Truly See the Diagrams in Visual Math Problems?
Related papers
- Exploring Solution Divergence and Its Effect on Large Language Model Problem Solving
- Ishigaki-IDS-Bench: A Benchmark for Generating Information Delivery Specification from BIM Information Requirements
- Subliminal Steering: Stronger Encoding of Hidden Signals
- MedStruct-S: A Benchmark for Key Discovery, Key-Conditioned QA and Semi-Structured Extraction from OCR Clinical Reports
- The End of Transformers? On Challenging Attention and the Rise of Sub-Quadratic Architectures
- Untangling the Mechanisms of Misleading Context in Medical Question Answering