1
Projector Is All You Train
投影器成训练核心？这篇论文挑战传统微调范式，或为高效训练提供新思路。
arXiv:2608.19726v1 Announce Type: cross Abstract: The typical training process of a multimodal large language model (MLLM) involves adapting both the …
2
MEUSLI: a Multilingual Projector for LLM-based ASR and Beyond
新研究MEUSLI：轻量多语言投影仪桥接语音编码器与LLM，提升ASR性能，值得关注。
arXiv:2607.22100v1 Announce Type: cross Abstract: Lightweight projectors are an established way to connect pre-trained speech encoders with large lang…
3
Query-based Cross-Modal Projector Bolstering Mamba Multimodal LLM
用Mamba替代Transformer，结合查询式跨模态投影器，大幅降低多模态大模型的计算复杂度
arXiv:2606.04719v1 Announce Type: new Abstract: The Transformer's quadratic complexity with input length imposes an unsustainable computational load o…
4
A More Word-like Image Tokenization for MLLMs
让图像分词更接近文本语义，提出新方法优化多模态大语言模型的融合效果。
arXiv:2605.17954v1 Announce Type: cross Abstract: Modern multimodal large language models (MLLMs) typically keep the language model fixed and train a …