5 papers
The MiniMax-M2 Series: Mini Activations Unleashing Max Real-World Intelligence
MiniMax, :, Aili Chen +219
We introduce the MiniMax-M2 series, a family of Mixture-of-Experts language models built around the principle that mini activations can unleash maximum real-world intelligence. The…
ExpertWeaver: Unlocking the Inherent MoE in Dense LLMs with GLU Activation Patterns
Ziyu Zhao, Tong Zhu, Zhi Zhang +6
Mixture-of-Experts (MoE) effectively scales model capacity while preserving computational efficiency through sparse expert activation. However, training high-quality MoEs from scra…
Train Less, Learn More: Adaptive Efficient Rollout Optimization for Group-Based Reinforcement Learning
Zhi Zhang, Zhen Han, Costas Mavromatis +9
Reinforcement learning (RL) plays a central role in large language model (LLM) post-training. Among existing approaches, Group Relative Policy Optimization (GRPO) is widely used, e…
Seed1.5-Thinking: Advancing Superb Reasoning Models with Reinforcement Learning
ByteDance Seed, :, Jiaze Chen +267
We introduce Seed1.5-Thinking, capable of reasoning through thinking before responding, resulting in improved performance on a wide range of benchmarks. Seed1.5-Thinking achieves 8…
Just Say What You Want: Only-prompting Self-rewarding Online Preference Optimization
Ruijie Xu, Zhihan Liu, Yongfei Liu +4
We address the challenge of online Reinforcement Learning from Human Feedback (RLHF) with a focus on self-rewarding alignment methods. In online RLHF, obtaining feedback requires i…