MiniMax Sparse Attention (MSA): a Two-Branch Block-Sparse Attention Trained on a 109B-Parameter MoE With a 3T-Token Budget
6月 17, 2026admin NUAI,Committee,ニュース,Uncategorized(0)MiniMax released MSA (MiniMax Sparse Attention), a sparse attention method built directly on Grouped Query...
MiniMax Releases MiniMax-Music3: An Open-Weights Music Model Generating Complete Five-Minute Songs From Lyrics and a Structured Caption
8月 17, 2026admin NUAI,Committee,ニュース,Uncategorized(0)MiniMax released MiniMax-Music3, an open-weights text-to-music model. The model takes two separate inputs: lyrics carrying...
MiniMax Releases MiniMax H3: An Omni-Modal Video Model That Generates 15-Second 2K Clips With Native Stereo Audio
8月 1, 2026admin NUAI,Committee,ニュース,Uncategorized(0)MiniMax releases MiniMax H3, a general-purpose multimodal generation model. MiniMax H3 is not a text-to-video...
MiniMax Releases M2.1: An Enhanced M2 Version with Features like Multi-Coding Language Support, API Integration, and Improved Tools for Structured Coding
12月 26, 2025admin NUAI,Committee,ニュース,Uncategorized(0)Just months after releasing M2—a fast, low-cost model designed for agents and code—MiniMax has introduced...
MiniMax AI Releases MiniMax-M1: A 456B Parameter Hybrid Model for Long-Context and Reinforcement Learning RL Tasks
6月 20, 2025admin NUAI,Committee,ニュース,Uncategorized(0)The Challenge of Long-Context Reasoning in AI Models Large reasoning models are not only designed...
Mind-altering substances are (still) falling short in clinical trials
3月 20, 2026admin NUAI,Committee,ニュース,Uncategorized(0)This week I want to look at where we are with psychedelics, the mind-altering substances...
Mind Your Moras: Orthography-Aware Error Analysis of Neural Japanese Morphological Generation
5月 25, 2026admin NUAI,Committee,ニュース,Uncategorized(0)arXiv:2605.20043v2 Announce Type: replace Abstract: We present an orthography-aware error analysis of Japanese past-tense morphological...
Mind the Gap: A Review of Arabic Post-Training Datasets and Their Limitations
7月 22, 2025admin NUAI,Committee,ニュース,Uncategorized(0)arXiv:2507.14688v1 Announce Type: new Abstract: Post-training has emerged as a crucial technique for aligning pre-trained...
Middo: Model-Informed Dynamic Data Optimization for Enhanced LLM Fine-Tuning via Closed-Loop Learning
9月 1, 2025admin NUAI,Committee,ニュース,Uncategorized(0)arXiv:2508.21589v1 Announce Type: new Abstract: Supervised Fine-Tuning (SFT) Large Language Models (LLM) fundamentally rely on...
