SGLang Supports Diffusion LLMs: Day-0 Implementation of LLaDA 2.0
Analysis
Key Takeaways
“SGLangにDiffusion LLM(dLLM)フレームワークを実装”
“SGLangにDiffusion LLM(dLLM)フレームワークを実装”
“GaMO expands the field of view from existing camera poses, which inherently preserves geometric consistency while providing broader scene coverage.”
“Generative classifiers...can avoid this issue by modeling all features, both core and spurious, instead of mainly spurious ones.”
“AODDiff inherently enables uncertainty quantification via multiple sampling, offering critical confidence metrics for downstream applications.”
“MDiffFR employs a tailored diffusion model on the server to generate embeddings for new items, which are then distributed to clients for cold-start inference.”
“ADS drives decoder success rates to near zero with minimal perceptual impact.”
“The paper's core finding is the ability to generate a high-quality, contextually grounded 3D mesh from a single RGB-D image in under one second.”
“SeedProteo achieves state-of-the-art performance among open-source methods, attaining the highest in-silico design success rates, structural diversity and novelty.”
“DiffThinker significantly outperforms leading closed source models including GPT-5 (+314.2%) and Gemini-3-Flash (+111.6%), as well as the fine-tuned Qwen3-VL-32B baseline (+39.0%), highlighting generative multimodal reasoning as a promising approach for vision-centric reasoning.”
“The model achieves comparable performance with less than 5% of the task-specific data required by dedicated expert models.”
“The framework incorporates a Salient Region Selection module and a Jacobian Vector Product Guidance mechanism to generate physically plausible adversarial objects.”
“The diffusion actions selected by deep Q-learning at different iterations indeed composite a stochastic anisotropic diffusion process with strong adaptivity to different image structures, which enjoys improvement over the traditional ones.”
“RoboPerform, the first unified audio-to-locomotion framework that can directly generate music-driven dance and speech-driven co-speech gestures from audio.”
“SGPS enables more accurate posterior sampling and reduces error accumulation, maintaining high reconstruction quality with fewer than 100 Neural Function Evaluations (NFEs).”
“The paper presents an alternating projected gradient descent and minimization algorithm for recovering a low-rank feature matrix in a diffusion-based decentralized and federated fashion.”
“SCPainter integrates 3D Gaussian Splat (GS) car asset representations and 3D scene point clouds with diffusion-based generation to jointly enable realistic 3D asset insertion and NVS.”
“The method achieves state-of-the-art performance in indoor benchmarks under constrained training conditions.”
““By explicitly constraining the generation with a goal image, our method enforces physical plausibility and goal consistency throughout the generated trajectory.””
“Dream-VLA achieves top-tier performance of 97.2% average success rate on LIBERO, 71.4% overall average on SimplerEnv-Bridge, and 60.5% overall average on SimplerEnv-Fractal, surpassing leading models such as $π_0$ and GR00T-N1.”
“The CCDDPM leverages the signal intensity-based 6G V2X Radio Environment Map (REM) from limited historical transmitter vehicles in a specific region, to predict the REMs for a transmitter vehicle with arbitrary coordinates across the same region.”
“DeFloMat achieves state-of-the-art accuracy ($43.32\% ext{ } AP_{10:50}$) in only $3$ inference steps, which represents a $1.4 imes$ performance improvement over DiffusionDet's maximum converged performance ($31.03\% ext{ } AP_{10:50}$ at $4$ steps).”
“Diffusion models offer a flexible framework for SBI tasks, addressing pain points of normalizing flows and offering robustness in non-ideal data conditions.”
“The paper proposes a two-stage autoregressive adaptation and acceleration framework to adapt a high-fidelity human video diffusion model for real-time, interactive streaming.”
“SpotEdit achieves efficient and precise image editing by reducing unnecessary computation and maintaining high fidelity in unmodified areas.”
“CrownGen surpasses state-of-the-art models in geometric fidelity and significantly reduces active design time.”
“SyncAnyone achieves state-of-the-art results in visual quality, temporal coherence, and identity preservation under in-the wild lip-syncing scenarios.”
“The bidirectional constraint makes visual predictions executable and keeps decisions grounded in physically consistent, task-relevant futures, mitigating cumulative errors common in decoupled 'envision-then-plan' pipelines.”
“”
“The paper introduces DiEC: Diffusion Embedded Clustering.”
“The paper focuses on rethinking the drafting strategy within speculative decoding.”
“The paper focuses on self-verified and efficient test-time scaling for diffusion multi-modal large language models.”
“The paper is available on ArXiv.”
“The study investigates sampling hyperparameters within the context of diffusion-based super-resolution.”
“DESSERT is a diffusion-based model.”
“Mitty is a diffusion-based human-to-robot video generation model.”
“The paper focuses on Informative Noise Enhanced Diffusion Based Contrastive Learning.”
“”
“”
“The research focuses on diffusion-based restoration for multi-modal 3D object detection.”
“SCAdapter is a method for content-style disentanglement in diffusion style transfer.”
“JiT (Just image Transformer) does not use VAE and performs flow-matching in pixel space. The model performs better by predicting the real image x (x-pred) rather than the velocity v.”
“Diffusion-Based Probabilistic Streamflow Forecasting with a State Space Backbone”
“”
“The article focuses on reducing the domain gap.”
“REST utilizes ID-Context Caching and Asynchronous Streaming Distillation.”
“”
“The article likely details the technical aspects of the diffusion model implementation and compares its performance against Perlin Noise.”
“The article focuses on dLLMs and early diffusion inference termination.”
“The research leverages a 300K-scale dataset.”
“The article likely explores the performance of this approach in terms of latency, cost, and overall system efficiency.”
Daily digest of the most important AI developments
No spam. Unsubscribe anytime.
Support free AI news
Support Us