Wave Forcing: Towards Speed-of-Light Streaming Video Generation
Yijie Jin 외
Wave Forcing enables efficient continuous long-video generation with consistent appearance and natural motion through co-designed generation, training, and execution, achieving high throughput and quality.
Video Generation Meets Agent Harness: Agentic Video Generation, Agentic World Modeling and Recursive Self-Improvement
Liaoyuan Fan 외
Agent harnesses integrate video generation and world models to enable planning, tool use, and recursive self-improvement for autonomous multi-shot video production.
Democratizing Robot-Use Agents: A Unified Harness with Open Models
Jiarui Yuan 외
Uni-robot-agent enables open-weight robot-use agents to surpass proprietary models in zero-shot manipulation and suggests joint development of four core capabilities for future training.
Follow the Gradient: Tracing Safety Degradation Back to Training Data
Saket Reddy 외
SimpleLoRIF uses gradient alignment to filter instruction data that silently degrades LLM safety, yielding a safer dataset that reduces adversarial success while preserving capabilities.
OSWAM: An Effective World Action Model with One-Step Action Generation
Guowen Zhang 외
OSWAM enables real-time, high-quality action generation in autonomous driving and embodied control by initializing diffusion with structured imagination tokens and distilling future-conditioned privileged information into a single denoising step.
From Fragmentation to Fusion: A Survey and Roadmap for Omnimodal Embeddings
Ziyin Zhang 외
Embedding research across modalities is converging toward shared principles via contrastive learning and Transformers, and a framework of coverage, comparability, and compositionality is proposed to unify it through transformation, alignment, and joint embedding.
NorthCUA Technical Report: Turbocharging Computer-Use Agents through Experience at Scale
Chaoyun Zhang 외
NorthCUA scales computer-use agents via an experience flywheel, recursive self-improvement, staged learning, and reinforcement learning to achieve state-of-the-art open-source GUI and web interaction results.
CrossSWE: Benchmarking Cross-Repository Joint Development for Coding Agents
Silin Chen 외
CrossSWE introduces a multi-repository benchmark showing that current LLMs struggle with cross-repository consistency during joint software development.
Wave Forcing: Towards Speed-of-Light Streaming Video Generation
Yijie Jin 외
Wave Forcing enables efficient continuous long-video generation with consistent appearance and natural motion through co-designed generation, training, and execution, achieving high throughput and quality.
Video Generation Meets Agent Harness: Agentic Video Generation, Agentic World Modeling and Recursive Self-Improvement
Liaoyuan Fan 외
Agent harnesses integrate video generation and world models to enable planning, tool use, and recursive self-improvement for autonomous multi-shot video production.
Democratizing Robot-Use Agents: A Unified Harness with Open Models
Jiarui Yuan 외
Uni-robot-agent enables open-weight robot-use agents to surpass proprietary models in zero-shot manipulation and suggests joint development of four core capabilities for future training.
Follow the Gradient: Tracing Safety Degradation Back to Training Data
Saket Reddy 외
SimpleLoRIF uses gradient alignment to filter instruction data that silently degrades LLM safety, yielding a safer dataset that reduces adversarial success while preserving capabilities.
OSWAM: An Effective World Action Model with One-Step Action Generation
Guowen Zhang 외
OSWAM enables real-time, high-quality action generation in autonomous driving and embodied control by initializing diffusion with structured imagination tokens and distilling future-conditioned privileged information into a single denoising step.
From Fragmentation to Fusion: A Survey and Roadmap for Omnimodal Embeddings
Ziyin Zhang 외
Embedding research across modalities is converging toward shared principles via contrastive learning and Transformers, and a framework of coverage, comparability, and compositionality is proposed to unify it through transformation, alignment, and joint embedding.
NorthCUA Technical Report: Turbocharging Computer-Use Agents through Experience at Scale
Chaoyun Zhang 외
NorthCUA scales computer-use agents via an experience flywheel, recursive self-improvement, staged learning, and reinforcement learning to achieve state-of-the-art open-source GUI and web interaction results.
CrossSWE: Benchmarking Cross-Repository Joint Development for Coding Agents
Silin Chen 외
CrossSWE introduces a multi-repository benchmark showing that current LLMs struggle with cross-repository consistency during joint software development.