TurboDiffusion: Accelerating Video Diffusion Models by 100-200 Times research paper by UC Berkeley, 2025

UC Berkeley · Multimodal and robotics · Dec 18, 2025 · 37 citations · 95 upvotes · unverified

TurboDiffusion accelerates video generation by 100-200x using attention acceleration, step distillation, and quantization, while maintaining video quality.

By Jintao Zhang, Kaiwen Zheng, Kai Jiang and 5 more · arXiv 2512.16093 · PDF · Code

UnverifiedHugging Face's summary; not yet checked by hand.

More from UC Berkeley

All 14

ArchitecturesUnverifiedOct 14, 2026Correction Training for Diffusion ModelsThis method adds a correction step to diffusion models to fix errors that accumulate during the image generation process.

UC Berkeley42 upvotesRead →Multimodal and roboticsUnverifiedSep 29, 2026OmniTaskonomy: When Does Visual Generation Improve Visual Understanding?This paper maps how visual generation tasks improve visual understanding capabilities, revealing specific task-dependent transfer benefits.

UC Berkeley57 upvotesRead →Inference and efficiencyUnverifiedAug 17, 2026FreeToken: Efficient Edge-Native MoE Serving with Bandwidth-Adaptive ExecutionFreeToken is an edge-native Mixture-of-Experts serving system that dynamically maps computation and model state onto heterogeneous local hardware to run large open-weight models on personal machines.

UC Berkeley8 citations86 upvotesRead →Agents and evaluationUnverifiedJun 17, 2026Playful Agentic Robot LearningEmbodied robots learn reusable skills through self-directed play and exploration, then apply these skills to improve performance on downstream tasks without additional training.

UC Berkeley37 citations49 upvotesRead →Agents and evaluationJun 3, 2026Agents' Last ExamOver 1,000 real, checkable professional tasks written with 250+ industry experts; the hardest tier is far from solved.

UC Berkeley42 citations393 upvotesRead →Inference and efficiencyUnverifiedMar 10, 2026Flash-KMeans: Fast and Memory-Efficient Exact K-MeansFlash-kmeans enables efficient online k-means clustering on GPUs through novel kernel-level optimizations that eliminate I/O bottlenecks and reduce atomic write contention.

UC Berkeley16 citations64 upvotesRead →Foundation modelsUnverifiedFeb 26, 2026dLLM: Simple Diffusion Language ModelingA unified open-source framework is presented that standardizes core components of diffusion language modeling for reproduction, customization, and accessible development of both large and small models.

UC Berkeley18 citations136 upvotesRead →Inference and efficiencyUnverifiedFeb 13, 2026SLA2: Sparse-Linear Attention with Learnable Routing and QATSLA2 improves sparse-linear attention in diffusion models by introducing a learnable router, direct attention formulation, and quantization-aware fine-tuning for enhanced efficiency and quality.

UC Berkeley1 citations59 upvotesRead →

Topic

Multimodal and roboticsApr 5, 2023Segment AnythingA promptable model and a dataset of over a billion masks that cut out any object in any image.

Meta16.8k citations6 upvotesRead →Multimodal and roboticsJan 30, 2023BLIP-2: Bootstrapping Language-Image Pre-training with Frozen Image Encoders and Large Language ModelsBLIP-2 efficiently pre-trains vision-language models using off-the-shelf frozen encoders and decoders, achieving state-of-the-art performance with fewer parameters.

Salesforce9,805 citations3 upvotesRead →Multimodal and roboticsJan 28, 2022BLIP: Bootstrapping Language-Image Pre-training for Unified Vision-Language Understanding and GenerationBLIP, a Vision-Language Pre-training framework, improves performance across both understanding and generation tasks by bootstrapping captions from noisy web data, achieving state-of-the-art results on image-text retrieval, image captioning, and VQA, and showing strong generalization to video-language tasks.

Salesforce7,655 citations5 upvotesRead →Multimodal and roboticsUnverifiedOct 25, 2024GPT-4o System CardGPT-4o is an omnimodal autoregressive model trained to handle text, audio, image, and video inputs, offering high-performance outputs across these modalities, with particular strengths in vision and audio.

OpenAI5,388 citations88 upvotesRead →Multimodal and roboticsUnverifiedNov 20, 2025SAM 3: Segment Anything with ConceptsSegment Anything Model 3 achieves state-of-the-art performance in promptable concept segmentation and tracking by leveraging a unified model architecture with decoupled recognition and localization.

Meta1,466 citations138 upvotesRead →Multimodal and roboticsUnverifiedMar 8, 2024DeepSeek-VL: Towards Real-World Vision-Language UnderstandingDeepSeek-VL is an open-source vision-language model that achieves state-of-the-art performance in real-world applications by combining a hybrid vision encoder with effective pretraining strategies to preserve language model capabilities.

DeepSeek939 citations49 upvotesRead →

About this paper

Venue: arXiv.org
Citations: 37, 5 influential · Semantic Scholar
Upvotes: 95 · Hugging Face
Lab: UC Berkeley

Changes

CitationsNov 2, 2026Citations rose from 33 to 37.

CitationsSep 28, 2026Influential citations rose from 4 to 5.

CitationsSep 28, 2026Citations rose from 32 to 33.

CitationsSep 25, 2026A first count of 4 influential citations.

CitationsSep 25, 2026A first count of 32 citations.

New paperSep 25, 2026Found by the weekly Hugging Face scan; not yet checked by hand.

What it shows

More from UC Berkeley

Other multimodal and robotics papers

Changes