【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:
[] 🧭 LongHorizon-Harness: Advancing Long-Horizon Agents for Real-World Tasks(LongHorizon-Harness:推动面向真实世界任务的长时程智能体)
[] 🎙 SwanTale: Unified Multi-Speaker Speech and Audio Generation for Instruct and Zero-Shot Tasks(SwanTale:面向指令与零样本任务的统一多说话人语音与音频生成)
[] 🎯 VAD: Attributing Visual Evidence for Target Reconstruction in Multimodal On-Policy Distillation(VAD:在多模态在策略蒸馏中为目标重建归因视觉证据)
[] 🤖 Progressive Agent Skill Generation via Reinforcement Learning(基于强化学习的渐进式智能体技能生成)
[] ⚓ DAPD: Dual-Anchored Policy Distillation(双锚定策略蒸馏)
[] 🧲 UEmbed: Unified Sparse and Dense Multimodal Embeddings(UEmbed:统一稀疏与稠密多模态嵌入)
[] 🌍 WorldExam: Benchmarking World Models from Apparent Appearance to Inherent Reactivity(WorldExam:从表象外观到内在反应性的世界模型基准评测)
[] 🔗 CADENA: Stepwise CAD Reverse Engineering(CADENA:逐步式CAD逆向工程)
[] 🛠 SKT: Skill-Use Training at Scale via Verified Synthetic Data Generation(SKT:通过经验证的合成数据生成实现规模化技能使用训练)
[] 🤖 SWE-Touch: Benchmarking Coding Agents When Users Touch the Code(SWE-Touch:在用户改动代码时对编码代理的基准测试)
[] 🚗 Deferred Exposure of Future Trajectories for Verifiable Reasoning in Autonomous Driving VLMs(自动驾驶视觉语言模型中用于可验证推理的未来轨迹延迟暴露)
[] 🧠 WCM: A World Critic Model for Vision-Language-Action Reinforcement Learning(WCM:面向视觉-语言-动作强化学习的世界评论家模型)
[] 🔄 Motion Beyond Morphology: Bootstrapping Cross-Category Motion Transfer from Abstract Motion Representations(超越形态的运动:从抽象运动表征引导跨类别运动迁移)
[] 🧠 GradCuit: Credit-Assigned Gradient Flow Enables Robust and Interpretable Test-Time Latent Reasoning(GradCuit:信用分配的梯度流实现稳健且可解释的测试时潜在推理)
[] 🛋 Roomer: Reflective Object-Grounded Model Editing and Repair for 3D Indoor Layout Synthesis(Roomer:面向三维室内布局合成的反思式对象级模型编辑与修复)

【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
