【赞助商】
OpenClaw快报
每天五分钟,听听 OpenClaw 快报,带你了解最新动态和业内讨论
传送门 www.xiaoyuzhoufm.com
【目录】
本期的 15 篇论文如下:
[] 🧠 Continual Learning Mechanisms Compose for Long-Horizon Memorization(持续学习机制组合用于长时程记忆)
[] 🎮 AI for Games in the Foundation Model Era(基础模型时代的游戏人工智能)
[] 🎙 StepAudio 3 Realtime Technical Report(StepAudio 3 Realtime 技术报告)
[] 🎵 StepAudio 3 Music Technical Report(StepAudio 3 Music 技术报告)
[] 🤖 ScienceBuddy: Recursive-in-Recursive Self-Improvement for Interactive Scientific Agents(ScienceBuddy:面向交互式科学智能体的递归嵌套式自我改进)
[] 🧭 HarnessVLN: Unifying Training-Free Embodied Navigation through an Agent Harness(HarnessVLN:通过智能体框架统一免训练具身导航)
[] 🤖 The Last AI Built by Humans: Toward Genuine Recursive Self-Improvement(人类建造的最后一个AI:迈向真正的递归自我改进)
[] 🏗 Another Blueprint In The Wall: How to Ask Frontier AI Like a Kid?(墙里的另一张蓝图:如何像孩子一样向前沿AI提问?)
[] 🧠 Mind2Dialogue: Training Human-Aware Language Models by Simulating User Mental States(Mind2Dialogue:通过模拟用户心理状态训练人类感知语言模型)
[] 🧩 ModularRSI: Modular and Generalizable Recursive Harness Self-Improvement(ModularRSI:模块化且可泛化的递归式执行框架自我改进)
[] 🤖 Modality-Autoregressive World-Action Models(模态自回归世界动作模型)
[] 📐 Disentangling Representation Evolution in Transformers through Directional Decomposition(通过方向分解解耦 Transformer 中的表征演化)
[] 🎥 PhysStream: Streaming Physics-Grounded Video Generation with Structured Scene Memory and Fine-Grained Motion Control(PhysStream:具备结构化场景记忆与细粒度运动控制的流式物理基础视频生成)
[] 🧪 ImpossibleRubrics: Stress-Testing Generated Rubrics as Reward Signals(ImpossibleRubrics:对作为奖励信号生成的评分标准进行压力测试)
[] 🧠 Convergent Emergence of In-Context Learning Across Modalities(跨模态上下文学习的趋同涌现)

【关注我们】
您还可以在以下平台找到我们,获得播客内容以外更多信息
小红书: AI速递
