ECoRAG: Evidentiality-guided Compression for Long Context RAG
arXiv:2506.05167v2 Announce Type: replace-cross Abstract: Large Language Models (LLMs) have shown remarkable performance in Open-Domain Question Answering (ODQA) by leveraging…
Optimization-Free Universal Watermark Forgery with Regenerative Diffusion Models
arXiv:2506.06018v1 Announce Type: cross Abstract: Watermarking becomes one of the pivotal solutions to trace and verify the origin of synthetic…
When to Trust Context: Self-Reflective Debates for Context Reliability
arXiv:2506.06020v1 Announce Type: cross Abstract: Large language models frequently encounter conflicts between their parametric knowledge and contextual input, often resulting…
Exploring Diffusion Transformer Designs via Grafting
arXiv:2506.05340v2 Announce Type: replace-cross Abstract: Designing model architectures requires decisions such as selecting operators (e.g., attention, convolution) and configurations (e.g.,…
A Path to Loving
arXiv:2506.05352v1 Announce Type: new Abstract: This work lays the foundations for a rigorous ontological characterization of love, addressing its philosophical…
Confidence-Guided Human-AI Collaboration: Reinforcement Learning with Distributional Proxy Value Propagation for Autonomous Driving
arXiv:2506.03568v2 Announce Type: replace-cross Abstract: Autonomous driving promises significant advancements in mobility, road safety and traffic efficiency, yet reinforcement learning…
ProRefine: Inference-time Prompt Refinement with Textual Feedback
arXiv:2506.05305v1 Announce Type: cross Abstract: Agentic workflows, where multiple AI agents collaborate to accomplish complex tasks like reasoning or planning,…