Tag: multimodal models
Browse our exclusive articles!
Stabilizing MLLM Self-Evolution with Softened Retracing
Enhance Multimodal LLMs self-evolution using Continuous Softened Retracing reSampling for stable, unbiased, and effective unsupervised learning.
UI-AGILE: Boost GUI Agents with RL & Precise Grounding
Discover UI-AGILE, a breakthrough in GUI agents using reinforcement learning and precise inference grounding for improved accuracy and performance.
V-Reflection: Enhancing MLLMs with Active Visual Reasoning
Discover how V-Reflection transforms MLLMs from passive observers to active interrogators, boosting accuracy and interpretability in visual tasks.
Hallucination’s Impact on Reinforcement Training in Multimodal Models
Explore how hallucination influences reinforcement post-training in multimodal reasoning models, boosting visual and multimodal AI performance.
Federated Pretraining for Multimodal Large Language Models
Discover Fed-CMP, a novel federated pretraining method enhancing multimodal large language models while preserving data privacy and improving performance.