ViQ Speeds Up Multimodal Training 20-70%

Featured

• ViQ: Text-Aligned Visual Quantized Representations at Any Resolution score 13
  入选 HF Daily Papers; HF 热度: 37 upvotes (+4); 有代码实现; 关键词(2): quantization, pre-training; 顶会接收: ECCV
• OPID: On-Policy Skill Distillation for Agentic Reinforcement Learning score 10
  入选 HF Daily Papers; HF 热度: 40 upvotes (+4); 有代码实现; 关键词(2): distillation, agentic
• DanceOPD: On-Policy Generative Field Distillation score 8
  入选 HF Daily Papers; HF 热度: 63 upvotes (+4); 关键词(3): distillation, state space, text-to-image
• Confidence-Aware Tool Orchestration for Robust Video Understanding score 8
  入选 HF Daily Papers; HF 热度: 9 upvotes (+2); 有代码实现; 关键词(5): GRPO, agentic, reasoning, embodied, open-source
• Qwen-Image-Agent: Bridging the Context Gap in Real-World Image Generation score 8
  入选 HF Daily Papers; HF 热度: 39 upvotes (+4); 关键词(2): agentic, text-to-image

Also Worth Noting

• Nemotron-TwoTower: Diffusion Language Modeling with Pretrained Autoregressive Context score 4
  机构: NVIDIA; 关键词(3): throughput, MoE, mamba
• From Hallucination to Grounding: Diagnosing Visual Spatial Intelligence via CRISP score 4
  关键词(3): post-training, reasoning, open-source; 顶会接收: ECCV
• PhyEditBench: A Real-World Multi-Stage Benchmark for Physics-Aware Image Editing score 4
  关键词(2): scaling, reasoning; 顶会接收: ECCV
• CAT-Q: Cost-efficient and Accurate Ternary Quantization for LLMs score 4
  关键词(2): quantization, post-training; 顶会接收: ICML
• LiveEdit: Towards Real-Time Diffusion-Based Streaming Video Editing score 4
  关键词(4): distillation, deployment, latency, real-time; 顶会接收: ECCV
• ProtoKV: Streaming Video Understanding under Delayed Query with Summary-State Memory score 4
  关键词(1): latency; 顶会接收: ICML
• ResilPhase: Plug-and-Play Phase Mapping and Noise-Resilient Macro-Trajectory Extrapolation for Diffusion Acceleration score 4
  关键词(1): latency; 顶会接收: ECCV
• LearniBridge: Learnable Calibration of Feature Caching for Diffusion Models Acceleration score 4
  关键词(1): lightweight; 顶会接收: ICML
• Reasoning Quality Emerges Early: Data Curation for Reasoning Models score 4
  关键词(3): fine-tuning, reasoning, data curation; 顶会接收: ICML
• PhysRAG: Enhancing Physics-Awareness in Video Generation via Retrieval-Augmented Generation score 4
  关键词(2): retrieval-augmented, RAG; 顶会接收: ECCV
• E-TTS: A New Embodied Test-Time Scaling Framework for Robotic Manipulation score 4
  关键词(4): scaling, reasoning, vision-language, embodied; 顶会接收: ECCV
• Empowering GUI Agents via Autonomous Experience Exploration and Hindsight Experience Utilization for Task Planning score 4
  关键词(1): open source; 顶会接收: ACL
• PhysiFormer: Learning to Simulate Mechanics in World Space score 6
  入选 HF Daily Papers; HF 热度: 9 upvotes (+2); 关键词(2): reasoning, robotics
• Don't Settle at the Mode! Mitigating Diversity Collapse in Pretrained Flow Models via Feature Self-Guidance score 4
  关键词(1): text-to-image; 顶会接收: ECCV
• Radical AI Interpretability score 3
  机构: Cambridge