Benchmark evaluations, applications, and challenges of large vision language models: A survey

Z Li, X Wu, H Du, H Nghiem, G Shi - arxiv preprint arxiv:2501.02189, 2025‏ - arxiv.org
Multimodal Vision Language Models (VLMs) have emerged as a transformative technology
at the intersection of computer vision and natural language processing, enabling machines …

Self-exploring language models: Active preference elicitation for online alignment

S Zhang, D Yu, H Sharma, H Zhong, Z Liu… - arxiv preprint arxiv …, 2024‏ - arxiv.org
Preference optimization, particularly through Reinforcement Learning from Human
Feedback (RLHF), has achieved significant success in aligning Large Language Models …

Llava-critic: Learning to evaluate multimodal models

T **ong, X Wang, D Guo, Q Ye, H Fan, Q Gu… - arxiv preprint arxiv …, 2024‏ - arxiv.org
We introduce LLaVA-Critic, the first open-source large multimodal model (LMM) designed as
a generalist evaluator to assess performance across a wide range of multimodal tasks …

Critic-v: Vlm critics help catch vlm errors in multimodal reasoning

D Zhang, J Lei, J Li, X Wang, Y Liu, Z Yang, J Li… - arxiv preprint arxiv …, 2024‏ - arxiv.org
Vision-language models~(VLMs) have shown remarkable advancements in multimodal
reasoning tasks. However, they still often generate inaccurate or irrelevant responses due to …

Llm uncertainty quantification through directional entailment graph and claim level response augmentation

L Da, T Chen, L Cheng, H Wei - arxiv preprint arxiv:2407.00994, 2024‏ - arxiv.org
The Large language models (LLMs) have showcased superior capabilities in sophisticated
tasks across various domains, stemming from basic question-answer (QA), they are …

Teaching-assistant-in-the-loop: Improving knowledge distillation from imperfect teacher models in low-budget scenarios

Y Zhou, W Ai - arxiv preprint arxiv:2406.05322, 2024‏ - arxiv.org
There is increasing interest in distilling task-specific knowledge from large language models
(LLM) to smaller student models. Nonetheless, LLM distillation presents a dual challenge: 1) …

Explore spurious correlations at the concept level in language models for text classification

Y Zhou, P Xu, X Liu, B An, W Ai, F Huang - arxiv preprint arxiv:2311.08648, 2023‏ - arxiv.org
Language models (LMs) have achieved notable success in numerous NLP tasks, employing
both fine-tuning and in-context learning (ICL) methods. While language models demonstrate …

Cream: Consistency regularized self-rewarding language models

Z Wang, W He, Z Liang, X Zhang, C Bansal… - arxiv preprint arxiv …, 2024‏ - arxiv.org
Recent self-rewarding large language models (LLM) have successfully applied LLM-as-a-
Judge to iteratively improve the alignment performance without the need of human …

Scaling inference-time search with vision value model for improved visual comprehension

W **yao, Y Zhengyuan, L Linjie, L Hong**… - arxiv preprint arxiv …, 2024‏ - arxiv.org
Despite significant advancements in vision-language models (VLMs), there lacks effective
approaches to enhance response quality by scaling inference-time computation. This …

Multi-stage balanced distillation: Addressing long-tail challenges in sequence-level knowledge distillation

Y Zhou, J Zhu, P Xu, X Liu, X Wang, D Koutra… - arxiv preprint arxiv …, 2024‏ - arxiv.org
Large language models (LLMs) have significantly advanced various natural language
processing tasks, but deploying them remains computationally expensive. Knowledge …