Tag: Meta-Learning
All the articles with the tag "Meta-Learning".
-
ReMA: Learning to Meta-think for LLMs with Multi-Agent Reinforcement Learning
Published: at 11:53 AM
85.15 🤔ReMA通过多智能体强化学习分离元思考和推理过程，提升了大型语言模型在数学推理和LLM-as-a-Judge任务上的性能，尤其在分布外泛化能力上表现出色，但对超参数敏感且多轮设置存在稳定性挑战。
-
Deciphering Trajectory-Aided LLM Reasoning: An Optimization Perspective
Published: at 11:20 AM 87.20 🤔本文提出RaML框架，从元学习视角将LLM推理轨迹视为伪梯度更新，通过理论分析和实验验证了推理与优化的关联，并探索了训练策略和轨迹特性对推理能力的提升潜力。
-
Warm Up Before You Train: Unlocking General Reasoning in Resource-Constrained Settings
Published: at 11:17 AM 93.37 🤔本文提出了一种两阶段训练框架，通过领域无关的Knights & Knaves逻辑游戏预热激活通用推理能力，并结合少量目标领域数据的RLVR训练，在资源受限环境下显著提升大型语言模型的推理性能和跨领域泛化能力。
-
Beyond Single-Task: Robust Multi-Task Length Generalization for LLMs
Published: at 11:14 AM 90.65 🤔本文提出Meta-RFFT框架，通过多任务规则跟随预训练和少量下游适应，显著提升了大型语言模型在未见任务上的长度泛化能力，32B模型在长度30的加法任务上达到98%准确率，超越现有长链推理模型。
-
Rethinking Meta-Learning from a Learning Lens
Published: at 11:18 AM 76.64 🤔This paper rethinks meta-learning from a 'learning' lens, proposing TRLearner, a plug-and-play method that leverages task relations to calibrate optimization, demonstrating significant performance improvements across regression, classification, drug activity, pose prediction, and OOD generalization tasks.