XConf Cuts Costs Tenfold; Agora Logs 165 Replications

Featured

• Confidence Comes from Experience: Experiential Confidence Estimation from Reasoning to Agents score 13
  机构: Cambridge; 入选 HF Daily Papers; HF 热度: 54 upvotes (+4); 有代码实现; 关键词(3): deployment, coding, reasoning
• ScienceIDE: Turning World's Scientific Codebase into Agent Learnable Environments score 10
  入选 HF Daily Papers; HF 热度: 73 upvotes (+4); 有代码实现; 关键词(2): fine-tuning, reasoning
• Rethinking Critic Learning in PPO: Understanding and Mitigating Value Flattening score 10
  入选 HF Daily Papers; HF 热度: 62 upvotes (+4); 有代码实现; 关键词(2): PPO, state space
• Agora: Git as Shared Memory for Collective AutoResearch score 10
  入选 HF Daily Papers; HF 热度: 42 upvotes (+4); 有代码实现; 关键词(1): coding
• Gaze as Evidence for Common Grounding: A Cross-Corpus Analysis of MapTask and MUNDEX score 9
  入选 HF Daily Papers; HF 热度: 22 upvotes (+4); 有代码实现
• In-Context Robot Learning with VLM Agents score 9
  入选 HF Daily Papers; HF 热度: 16 upvotes (+3); 有代码实现; 关键词(4): deployment, agentic, vision-language, embodied
• EvolveTrade: Experience-Driven Policy Refinement for Self-Evolving LLM Trading Agents score 8
  入选 HF Daily Papers; HF 热度: 31 upvotes (+4); 关键词(2): deployment, tool use
• Zing-0.5: Toward Playable Worlds with Real-Time Joint Action and Text Control score 8
  入选 HF Daily Papers; HF 热度: 30 upvotes (+4); 关键词(3): distillation, serving, real-time
• A Zeroth-Order Paradigm for LLM Preference Alignment score 8
  入选 HF Daily Papers; HF 热度: 22 upvotes (+4); 关键词(1): fine-tuning
• PANORAMA: Panoptic Grounded Captioning via Mask Proposal Selection score 9
  入选 HF Daily Papers; HF 热度: 10 upvotes (+3); 有代码实现; 关键词(1): vision-language

Also Worth Noting

• Locating Hidden Failures Makes Long-Horizon Agents More Reliable score 4
  机构: Google; 关键词(1): deployment
• QuanText: Protecting Dataset-Level Secrets in Textual Data Sharing score 4
  机构: Microsoft Research; 关键词(1): quantization
• Learning Fractional-Order Dynamics from a Single Trajectory score 4
  机构: Carnegie Mellon; 关键词(1): scaling
• DualSQL: Text-to-SQL with Multi-Agent Reinforcement Learning score 4
  机构: Microsoft; 关键词(2): agentic, reasoning
• MoRE: Mixture of Reused Experts score 4
  机构: Carnegie Mellon; 关键词(2): lightweight, MoE
• Generalist-Specialist Mixture-of-Experts for Rare Pathology Detection in Multimodal Imaging score 4
  机构: Imperial College; 关键词(1): MoE
• Enhancing Extubation Failure Prediction with LLM-Derived Features from Respiratory Therapy Clinical Notes score 3
  机构: Allen Institute