Papers arxiv:2605.06376 Continuous-Time Distribution Matching for Few-Step Diffusion Distillation Published on May 7 Submitted by liutao on May 8 alibaba-inc Authors: , , , , Zhengrong Yue , Zihao Pan , , , , , Abstract Continuous-Time Distribution Matching migrates diffusion model distillation fro… …
Papers arxiv:2605.30581 Prior Availability in Industrial Visual Sim-to-Real: A Review of CAD-Guided and CAD-Unavailable Regimes Published on May 28 Submitted by chenxi on Jun 3 Authors: Chenxi Tao , Abstract Industrial visual sim-to-real is reframed as a domain-gap problem categorized by prior avai… …
… However, SFT introduces distributional drift that neither preserves the model's original capabilities nor faithfully matches the supervision distribution. …
Papers arxiv:2604.27085 Efficient Training on Multiple Consumer GPUs with RoundPipe Published on Apr 29 Submitted by Kewei Zhang on May 1 Authors: , , , , Abstract RoundPipe introduces a novel pipeline scheduling approach that eliminates weight binding constraints in LLM fine-tuning, enabling effic… …
… Although methods like Negative Sample Reinforcement NSR mitigate this issue by upweighting penalty from negative samples, they may suppress the semantic distributions shared between positive and negative responses. …
… Spatial Adaptive Multi Guidance in Diffusion Models 2026 Diffusion-APO: Trajectory-Aware Direct Preference Alignment for Video Diffusion Transformers 2026 Diffusion Image Generation with Explicit Modeling of Data Manifold Geometry 2026 Continuous-Time Distribution Matching for Few-Step Diffusion Di… …
… Domino first uses a parallel draft backbone to produce preliminary draft distributions for the entire block, and then applies a lightweight Domino head to refine them with prefix-dependent causal information. …
… Formally, we prove that constraining the learned field to be conservative reduces hypothesis complexity and tightens out-of-distribution generalization bounds. …
… AI-generated summary Domain Generalization DG aims to learn representations that remain robust under out-of-distribution OOD shifts and generalize effectively to unseen target domains. …
… However, existing Tree-RAG methods designed for single-document retrieval face critical challenges in scaling to cross-document multi-hop questions : 1 poor distribution adaptability, where k-means clustering introduces noise due to rigid distribution assumptions; 2 structural isolation, as tree in… …