Defense in Depth
ai safety
Defense in depth applied to AI safety is the principle of layering multiple independent safety mechanisms so that no single failure can lead to harmful outcomes. Borrowed…
997 technical terms and definitions
ai safety
Defense in depth applied to AI safety is the principle of layering multiple independent safety mechanisms so that no single failure can lead to harmful outcomes. Borrowed…
Defensive Distillation Introduction & Motivation Defensive Distillation addresses a central problem in robustness of advrobust models under adversarial conditions: how to Maintain…
Defensive Distillation (Extended Study) Introduction & Motivation Defensive Distillation (Extended Study) addresses a central problem in robustness of advrobust models under…
Defensive Distillation (Extended Study) Introduction & Motivation Defensive Distillation (Extended Study) addresses a central problem in robustness of advrobust models under…
defi decentralized finance, defi decentralized, decentralized finance
DeFi Decentralized Finance Introduction & Motivation DeFi Decentralized Finance addresses a central problem in blockchain process control and financial-primitive-automation…
Deformable Convolution Networks Introduction & Motivation Deformable Convolution Networks addresses a central problem in vision processing and adaptive-receptive-field prediction…
computer vision
Deformable models are 3D representations that can change shape through controlled deformations — enabling animation, shape matching, and morphing by defining how geometry…
deformable object manipulation, deformable object, object manipulation
Deformable Object Manipulation Introduction & Motivation Deformable Object Manipulation addresses a central problem in robotics process control and cloth-rope-deformable-material…
multimodal ai
Deformation Field is a learned mapping that warps coordinates between canonical and observed dynamic scene states - It enables motion-aware reconstruction in dynamic neural…
reliability
Degraded failure analysis is the failure analysis approach that studies parametric drift and partial-function degradation before catastrophic breakdown - it captures early warning…
deit, data-efficient image transformer, computer vision
DeiT - Data-Efficient Image Transformers Introduction & Motivation DeiT: data-efficient vision transformers via knowledge distillation. Train on ImageNet without large-scale…
deit data efficient image transformers, deit data, efficient image transformers, 21k ce cnn gpu kl v2, deit data-efficient image, data-efficient image transformers, deit data-efficient
DeiT - Data-Efficient Image Transformers Introduction & Motivation DeiT: data-efficient vision transformers via knowledge distillation. Train on ImageNet without large-scale…
deit data-efficient training and knowledge distillation prediction, deit data-efficient training, data-efficient training knowledge, training knowledge distillation
DeiT Data-Efficient Training and Knowledge Distillation Prediction Introduction & Motivation DeiT Data-Efficient Training and Knowledge Distillation Prediction addresses a central…
delaunay triangulation properties, delaunay triangulation, triangulation properties
Delaunay Triangulation Properties Introduction & Motivation Delaunay Triangulation Properties addresses a central problem in algorithms process control…
delay-locked loop lock range prediction, delay-locked loop lock, loop lock range, lock range prediction, delay-locked loop, loop lock, lock range, range prediction
Delay-Locked Loop Lock Range Prediction with Machine Learning Introduction & Motivation Delay-Locked Loop Lock Range Prediction with Machine Learning addresses a central problem…
ai safety
Delimiter-based protection is the prompt-hardening technique that uses explicit boundary markers to separate trusted instructions from untrusted input content - it improves…
Delta Method for Non-Linear Function Variance Propagation Introduction & Motivation Delta Method for Non-Linear Function Variance Propagation addresses a central problem…
delta-sigma adc modulation stability prediction, delta-sigma adc modulation, adc modulation stability, modulation stability prediction, delta-sigma adc, adc modulation
Delta-Sigma ADC Modulation Stability Prediction with Machine Learning Introduction & Motivation Delta-Sigma ADC Modulation Stability Prediction with Machine Learning addresses…
environmental & sustainability
Demand Control Ventilation is ventilation control that adjusts outside-air intake based on measured occupancy or air-quality indicators - It reduces unnecessary conditioning load…
supply chain & logistics
Demand Forecasting is prediction of future product demand to guide procurement, production, and inventory decisions - It aligns supply commitments with expected market needs. What…
advanced training
Democratic co-learning is a collaborative semi-supervised framework where multiple learners vote and share pseudo labels - Consensus-based labeling aggregates multiple model…
Democratic Co-Learning with Diverse Model Architectures Introduction & Motivation Democratic Co-Learning with Diverse Model Architectures addresses a central problem…
equal outcome, fair
Demographic Parity is the fairness constraint requiring that an AI model's positive prediction rate be equal across all demographic groups — one of the foundational fairness…
fairness
Demographic Parity is the fairness criterion requiring that an AI system's positive prediction rate be equal across all protected demographic groups — meaning that the probability…
Demography: Population Change, Fertility, Mortality, and Migration Demography studies population size, composition, distribution, and change. It measures births, deaths…
dendrite formation electroplating prediction, dendrite formation electroplating, formation electroplating prediction, dendrite formation, formation electroplating
Dendrite Formation Electroplating Prediction with Machine Learning Introduction & Motivation Dendrite Formation Electroplating Prediction with Machine Learning addresses a central…
diffusion, probabilistic, model, DDPM
Denoising Diffusion Probabilistic Models (DDPM) is a generative model class that iteratively denoises corrupted data samples over a series of diffusion steps — learning to reverse…
denoising autoencoder corruption robustness and noise injection control, denoising autoencoder corruption, autoencoder corruption robustness, corruption robustness noise
Denoising Autoencoder Corruption Robustness and Noise Injection Control Introduction & Motivation Denoising Autoencoder Corruption Robustness and Noise Injection Control addresses…
accelerated sampling diffusion, deterministic sampling, noise schedule diffusion, fast diffusion inference
Denoising Diffusion Implicit Models (DDIM) is a class of generative models that reformulate the diffusion sampling process as a non-Markovian deterministic mapping, enabling…
Denoising Diffusion Implicit Models (DDIM) Introduction & Motivation Denoising Diffusion Implicit Models (DDIM) addresses a central problem in generative modeling of diffusion…
Denoising Diffusion Implicit Models (DDIM) (Extended Study) Introduction & Motivation Denoising Diffusion Implicit Models (DDIM) (Extended Study) addresses a central problem…
Denoising Diffusion Implicit Models (DDIM) (Extended Study) Introduction & Motivation Denoising Diffusion Implicit Models (DDIM) (Extended Study) addresses a central problem…
Denoising Diffusion Implicit Models (DDIM) (Extended Study) Introduction & Motivation Denoising Diffusion Implicit Models (DDIM) (Extended Study) addresses a central problem…
denoising diffusion probabilistic models, ddpm, generative models
Denoising Diffusion Probabilistic Models (DDPMs) provide the core mathematical framework for diffusion-based generative models, learning to reverse a gradual noising process…
Denoising Diffusion Probabilistic Models (DDPM) Introduction & Motivation Denoising Diffusion Probabilistic Models (DDPM) addresses a central problem in generative modeling…
Denoising Diffusion Probabilistic Models (DDPM) (Extended Study) Introduction & Motivation Denoising Diffusion Probabilistic Models (DDPM) (Extended Study) addresses a central…
Denoising Diffusion Probabilistic Models (DDPM) (Extended Study) Introduction & Motivation Denoising Diffusion Probabilistic Models (DDPM) (Extended Study) addresses a central…
Denoising Diffusion Probabilistic Models (DDPM) (Extended Study) Introduction & Motivation Denoising Diffusion Probabilistic Models (DDPM) (Extended Study) addresses a central…
generative models
Denoising Score Matching (DSM) is a computationally efficient variant of score matching that estimates the score function ∇_x log p(x) by training a neural network to denoise…
Denoising Score Matching Introduction & Motivation Denoising Score Matching addresses a central problem in generative generation and score-function-accuracy synthesis: how…
generative models
Denoising strength is the parameter that controls the proportion of noise applied before reverse diffusion during conditional generation or editing - it sets the effective edit…
multimodal ai
Dense captioning is the task that detects multiple regions in an image and generates a descriptive caption for each region - it combines localization and language generation…
Dense Image Captioning by Regions Introduction & Motivation Dense Image Captioning by Regions addresses a central problem in multimodal learning and regional-caption-accuracy…
model architecture
Dense models activate all parameters for every input, the standard architecture for most neural networks. Definition: Every parameter participates in every forward pass. All…
Dense Object Captioning Introduction & Motivation Dense Object Captioning addresses a central problem in multimodal learning and object-caption-accuracy alignment: how to Learn…
bi encoder, dpr, embedding model, semantic search, sentence embedding retrieval
Dense Retrieval and Embedding Models are the neural information retrieval systems that encode queries and documents into dense vector representations in a shared semantic space…
Dense Video Captioning Introduction & Motivation Dense Video Captioning addresses a central problem in multimodal learning and temporal-localization-accuracy alignment: how…
neural architecture search
DenseNAS is NAS method emphasizing dense connectivity and width-aware architecture optimization. - It extends search beyond operator choice to include channel allocation…
densenet dense connection and feature reuse, densenet dense connection, dense connection feature, connection feature reuse, densenet dense, dense connection, connection feature
DenseNet Dense Connection and Feature Reuse Introduction & Motivation DenseNet Dense Connection and Feature Reuse addresses a central problem in DenseNet with dense inter-block…
Density-Peaks Clustering for Initial Active Seed Selection Introduction & Motivation Density-Peaks Clustering for Initial Active Seed Selection addresses a central problem…