Frontier Agents Finish One Task in Five at 1.6-Hour Length

Featured

• Bridging VideoQA and Video-Guided Agentic Tasks via Generalized Keyframe Extraction score 13
  入选 HF Daily Papers; HF 热度: 21 upvotes (+4); 有代码实现; 关键词(1): agentic; 顶会接收: ECCV
• OSWorld2.0: Benchmarking Computer Use Agents on Long-Horizon Real-World Tasks score 9
  入选 HF Daily Papers; HF 热度: 15 upvotes (+3); 有代码实现; 关键词(2): coding, reasoning
• PolicyGuard: A Dialogue-Grounded Sub-Agent Verifier for Policy Adherence in LLM Agents score 8
  入选 HF Daily Papers; HF 热度: 5 upvotes (+2); 有代码实现; 关键词(1): reasoning
• MirrorPPR: Exemplar-Based Portrait Photo Retouching score 8
  入选 HF Daily Papers; 有代码实现; 顶会接收: ECCV
• One Scene, Two Depths: Probing Geometric Ambiguity in Monocular Foundation Models score 5
  入选 HF Daily Papers; 有代码实现

Also Worth Noting

• Evidence-Informed LLM Beliefs for Continual Scientific Discovery score 4
  机构: Allen Institute; 关键词(2): retrieval-augmented, reasoning
• DTI: Dynamic Trajectory Initialization for Generative Face Video Super-Resolution score 4
  关键词(1): fine-tuning; 顶会接收: ECCV
• BrainRiem: Riemannian Prototype Learning for Source-Free Cross-Site Brain Network Diagnosis score 4
  关键词(1): serving; 顶会接收: ECCV
• Pointer-CAD v2: Plan-Then-Construct CAD Generation with Dimension-Aware Parametric Precision score 4
  关键词(4): quantization, production, edge, reasoning; 顶会接收: ECCV
• Multi-scale Object-Aware Gaze Estimation via Geometric Reasoning score 4
  关键词(1): reasoning; 顶会接收: ECCV
• When LLMs Develop Languages: Symbolic Communication for Efficient Multi-Agent Reasoning score 4
  机构: University of Toronto; 关键词(2): latency, reasoning
• NaLA: A 3D Native LLM Layout Agent for High-quality 3D Scene Generation score 4
  关键词(1): reasoning; 顶会接收: ECCV
• From Phase to Phenomenon: Self-Supervised Learning of Subsurface Scattering with Minimal Phase-shift Inputs score 4
  关键词(1): pretraining; 顶会接收: ECCV
• MIRROR: Aligning Semantic Relations from Language to Image via Gromov--Wasserstein score 4
  关键词(1): vision-language; 顶会接收: ECCV
• Harvesting AI Computation at the Edge via Generic Approximation score 4
  机构: Huawei; 关键词(1): edge
• Do Models Read What They Write? Causal Registers in Scratchpad Reasoning score 4
  机构: Stanford; 关键词(1): reasoning
• GarmentZoom: Generating Zoomable Images from Garment Listings score 4
  机构: University of Washington; 关键词(1): fine-tuning
• Coverage-Driven KV Cache Eviction for Efficient and Improved Inference of LLM score 4
  机构: Apple; 关键词(2): deployment, reasoning
• ScAle: Attention Head Scaling as a Minimal Adapter for Spatial Reasoning in Vision Language Models score 4
  关键词(4): scaling, lightweight, fine-tuning, reasoning; 顶会接收: ECCV
• Can Machines Really See Objects in Images? A Study Based on Syntactic Distance and Visual Self-Referential Instances score 3
  机构: Microsoft Research