Streaming Video QA Hits 2 FPS, RLVR Shrugs Off Noisy Labels

Featured

• AURA: Always-On Understanding and Real-Time Assistance via Video Streams score 10
  入选 HF Daily Papers; HF 热度: 39 upvotes (+4); 有代码实现; 关键词(2): deployment, real-time
• ClawArena: Benchmarking AI Agents in Evolving Information Environments score 10
  入选 HF Daily Papers; HF 热度: 27 upvotes (+4); 有代码实现; 关键词(1): reasoning
• Can LLMs Learn to Reason Robustly under Noisy Supervision? score 8
  入选 HF Daily Papers; HF 热度: 9 upvotes (+2); 有代码实现; 关键词(1): reasoning
• The Geometric Alignment Tax: Tokenization vs. Continuous Geometry in Scientific Foundation Models score 7
  入选 HF Daily Papers; HF 热度: 3 upvotes (+1); 有代码实现; 关键词(2): compression, quantization
• ACES: Who Tests the Tests? Leave-One-Out AUC Consistency for Code Generation score 5
  入选 HF Daily Papers; HF 热度: 4 upvotes (+1); 关键词(1): code generation

Also Worth Noting

• FactReview: Evidence-Grounded Reviews with Literature Positioning and Execution-Based Claim Verification score 4
  入选 HF Daily Papers; HF 热度: 3 upvotes (+1)
• Semantic IDs for Recommender Systems at Snapchat: Use Cases, Technical Challenges, and Design Choices score 4
  机构: MIT; 关键词(2): quantization, production
• GeoBrowse: A Geolocation Benchmark for Agentic Tool Use with Expert-Annotated Reasoning Traces score 4
  机构: Tencent; 关键词(4): agentic, tool use, reasoning, open-source
• Schema-Aware Planning and Hybrid Knowledge Toolset for Reliable Knowledge Graph Triple Verification score 4
  机构: Huawei; 关键词(1): reasoning
• Combee: Scaling Prompt Learning for Self-Improving Language Model Agents score 4
  机构: Stanford; 关键词(2): scaling, agentic
• 4C4D: 4 Camera 4D Gaussian Splatting score 3
  顶会接收: CVPR