timestep
BiDM: Pushing the Limit of Quantization for Diffusion Models
Diffusion models (DMs) have been significantly developed and widely used in various applications due to their excellent generative qualities.
- Research Report > New Finding (1.00)
- Research Report > Experimental Study (0.93)
LTMD: Learning Improvement of Spiking Neural Networks with Learnable Thresholding Neurons and Moderate Dropout
This limits SNNs' capability of transmitting precise information in the network, which becomes worse for deeper SNNs.
- Europe > Italy > Calabria > Catanzaro Province > Catanzaro (0.04)
- Asia > South Korea > Seoul > Seoul (0.04)
- Research Report > Experimental Study (1.00)
- Research Report > New Finding (0.93)
- Information Technology > Artificial Intelligence > Vision (1.00)
- Information Technology > Artificial Intelligence > Natural Language > Large Language Model (0.93)
- Information Technology > Artificial Intelligence > Machine Learning > Neural Networks > Deep Learning (0.93)
- Information Technology > Sensing and Signal Processing > Image Processing (0.92)
Understanding Hallucinations in Diffusion Models through Mode Interpolation
But where do such hallucinations come from?
- North America > United States > Pennsylvania > Allegheny County > Pittsburgh (0.04)
- Europe > Germany > Bavaria > Upper Bavaria > Munich (0.04)
- Asia (0.04)
- Research Report > Experimental Study (1.00)
- Research Report > New Finding (0.93)
Learning-to-Cache: Accelerating Diffusion Transformer via Layer Caching
Notably, for U-ViT -H/2 on ImageNet, approximately 93.68% of layers are cacheable in the cache step,
- Research Report > New Finding (1.00)
- Research Report > Experimental Study (0.93)
Superposed Decoding: Multiple Generations from a Single Autoregressive Inference Pass Ethan Shen Alan Fan Sarah Pratt Jae Sung Park Matthew Wallingford Sham Kakade Ari Holtzman Ranjay Krishna
Many applications today provide users with multiple auto-complete drafts as they type, including GitHub's code completion, Gmail's smart compose, and Apple's messaging auto-suggestions.
- Oceania > Australia > New South Wales (0.05)
- Europe > United Kingdom > England > Oxfordshire > Oxford (0.04)
- Asia > India > NCT > New Delhi (0.04)
- (3 more...)
- Education (0.94)
- Health & Medicine > Therapeutic Area (0.69)
- Information Technology > Services (0.46)
- Health & Medicine > Consumer Health (0.46)
- Research Report > Experimental Study (1.00)
- Research Report > New Finding (0.93)
Advancing Training Efficiency of Deep Spiking Neural Networks through Rate-based Backpropagation
Recent insights have revealed that rate-coding is a primary form of information representation captured by surrogate-gradient-based Backpropagation Through Time (BPTT) in training deep Spiking Neural Networks (SNNs).
- Europe > Netherlands > North Holland > Amsterdam (0.04)
- Asia > China > Zhejiang Province (0.04)
- Information Technology (0.67)
- Education (0.47)
Modeling Million-byte Sequences with Multiscale Transformers Lili Y u Dániel Simig Colin Flaherty Armen Aghajanyan Luke Zettlemoyer Mike Lewis Meta AI
Sequences of millions of bytes are ubiquitous; for example, music, image, or video files typically consist of multiple megabytes.
Remix-DiT: Mixing Diffusion Transformers for Multi-Expert Denoising Gongfan Fang Xinyin Ma Xinchao Wang National University of Singapore
The goal of multi-expert denoising is to increase the overall capacity of diffusion models while keeping an acceptable overhead.