Wan2.1-14B
PulseAugur coverage of Wan2.1-14B — every cluster mentioning Wan2.1-14B across labs, papers, and developer communities, ranked by signal.
3 day(s) with sentiment data
-
PISA method boosts Diffusion Transformer efficiency with piecewise sparse attention
Researchers have developed PISA, a new training-free method for Piecewise Sparse Attention designed to improve the efficiency of Diffusion Transformers. PISA addresses the quadratic complexity of attention in these mode…
-
SpecKCast accelerates diffusion transformers using frequency-domain Kalman filtering
Researchers have developed SpecKCast, a novel training-free method designed to accelerate diffusion transformers, which are commonly used for image and video generation. This technique forecasts the transformer's output…
-
VideoAWR method enhances video generation models with decoupled reinforcement learning
Researchers have introduced VideoAWR, a novel method for reinforcing video flow models that addresses a cold-start pathology in reinforcement learning for generative models. VideoAWR decouples the sampling process from …
-
Spectrum method accelerates diffusion model sampling with Chebyshev polynomials
Researchers have developed a novel training-free method called Spectrum to accelerate diffusion model sampling. This approach forecasts latent features at future diffusion steps by approximating them with Chebyshev poly…
-
RoLA attention mechanism boosts Diffusion Transformer efficiency for video generation
Researchers have developed RoLA, a novel attention mechanism designed to improve the efficiency of Diffusion Transformers (DiTs) used in video generation. This new method addresses the quadratic scaling issue of standar…
-
New PhyS framework distills physical priors into streaming world models
Researchers have developed PhyS, a novel three-stage framework designed to imbue streaming world models with physical coherence. This framework addresses limitations in current methods by constructing a large dataset of…
-
New methods LoSA and HEART accelerate video diffusion transformers
Researchers have developed two new methods, LoSA and HEART, to accelerate video diffusion transformers by optimizing sparse attention mechanisms. LoSA focuses on maintaining near-lossless fidelity by identifying and rem…
-
New RACER controller boosts diffusion model speed and reliability · 2 sources tracked
Researchers have developed RACER, a new closed-loop controller designed to improve the efficiency and reliability of diffusion models. Unlike previous methods that blindly trust forecasts, RACER analyzes the agreement b…
-
CachedSearch accelerates video diffusion model search with novel caching
Researchers have developed CachedSearch, a novel training-free method to accelerate test-time search for video diffusion models. This technique significantly reduces the computational cost of generating high-quality vid…
-
New DigenRL framework accelerates diffusion generative LLMs with disaggregated RL · 3 sources tracked
Researchers have developed DigenRL, a disaggregated reinforcement learning framework designed to enhance the efficiency of diffusion-based generative large language models. This new framework addresses limitations in ex…
-
New Steady-Forcing framework improves long-horizon nature video generation · 2 sources tracked
Researchers have developed Steady-Forcing, a new framework designed to improve the quality of long-horizon nature videos generated by autoregressive diffusion models. This method addresses the common issues of drifting …
-
PARE method enhances video generation efficiency with adaptive routing
Researchers have introduced PARE, a novel method for making Video Diffusion Transformers (DiTs) more computationally efficient. PARE addresses the high compute demands of DiTs by jointly compressing model width and dept…