compute-bound operations
model optimization
Compute-Bound Operations is operators whose speed is limited by arithmetic capacity rather than memory transfer - They benefit most from vectorization and accelerator-specific…
15,747 technical terms and definitions
model optimization
Compute-Bound Operations is operators whose speed is limited by arithmetic capacity rather than memory transfer - They benefit most from vectorization and accelerator-specific…
training
Compute-constrained regime is the training regime where available compute is the primary limiting factor on model and data scaling choices - it forces tradeoffs between model…
chip export controls, compute thresholds, 10^25 flop, sastry et al 2024, computing power and the governance of ai, training compute reporting, executive order 14110
Compute governance is the use of control over computing resources, especially the specialized chips and data centers used to train and run large AI models, as a lever for shaping…
compute-in-memory adc quantization error prediction, compute-in-memory adc quantization, adc quantization error, quantization error prediction, adc quantization
Compute-in-Memory ADC Quantization Error Prediction with Machine Learning Introduction & Motivation Compute-in-Memory ADC Quantization Error Prediction with Machine Learning…
model training
Scaling laws are the empirical power-law relationships that predict how a language model's loss falls as you add parameters, training data, and compute. They are the reason…
training
Scaling laws are the empirical power-law relationships that predict how a language model's loss falls as you add parameters, training data, and compute. They are the reason…
generative design, topology optimization, CAD copilot, AI chip floorplanning, EDA machine learning, Mirhoseini 2021, DSO.ai
AI in computer-aided design is the use of machine learning, optimization and generative models inside the software engineers use to design physical products and chips, so that…
computer vision, images, pixels, geometry, rendering, image processing, visual recognition, convolutional neural network
Computer graphics and vision are two directions on the exact same path: graphics turns numbers into an image, and vision turns an image back into numbers a program can reason…
authentication, authorization, cryptography, secure programming, threat modeling, least privilege, SQL injection, encryption
Computer security is what has to be added once the distributed systems from the previous topic turn a collection of trusting, communicating machines into something an attacker can…
image classification, object detection, image segmentation, vision transformer, wafer inspection
computer vision is the field that enables machines to extract structure, identity, geometry, motion, and meaning from images and video. It drives inspection, autonomy, robotics…
privacy, algorithmic fairness, intellectual property, social consequences, responsible computing
Computing ethics is the final and most important lens this series applies, because it asks not whether a system works, scales, or trains accurately, but whether it should exist…
tcav explainability, high-level concept testing, interpretability
TCAV (Testing with Concept Activation Vectors) is the high-level explainability method that tests how much a neural network relies on human-interpretable concepts — going beyond…
Concept Activation Vector Explanations Introduction & Motivation Concept Activation Vector Explanations addresses a central problem in explainability interpretation and model…
Concept Activation Vectors (TCAV) Introduction & Motivation Concept Activation Vectors (TCAV) addresses a central problem in interpretability and explanation of xai predictions…
Testing with Concept Activation Vectors TCAV Introduction & Motivation Testing with Concept Activation Vectors TCAV addresses a central problem in explainability interpretation…
Concept Activation Vectors (TCAV) (Extended Study) Introduction & Motivation Concept Activation Vectors (TCAV) (Extended Study) addresses a central problem in interpretability…
Concept Activation Vectors (TCAV) (Extended Study) Introduction & Motivation Concept Activation Vectors (TCAV) (Extended Study) addresses a central problem in interpretability…
explainable ai
Concept Bottleneck Models are neural network architectures that structure predictions through human-interpretable concepts as intermediate representations — forcing models…
Concept Bottleneck Models with Intervenable Concepts Introduction & Motivation Concept Bottleneck Models with Intervenable Concepts addresses a central problem in explainability…
concept drift detection, concept drift, drift detection
Concept Drift Detection Introduction & Motivation Concept Drift Detection addresses a central problem in timeseries process control and drift-definition-distribution-change…
Page-Hinkley Test for Sequential Concept Drift Introduction & Motivation Page-Hinkley Test for Sequential Concept Drift addresses a central problem in (subject) assurance…
Concept Importance Ranking Introduction & Motivation Concept Importance Ranking addresses a central problem in explainability interpretation and model understanding: how…
Concept Whitening for Inherently Disentangled Layers Introduction & Motivation Concept Whitening for Inherently Disentangled Layers addresses a central problem in explainability…
Concrete Dropout Introduction & Motivation Concrete Dropout addresses a central problem in bayesiandl modeling with calibrated predictive uncertainty: how to Produce…
Concrete Dropout (Extended Study) Introduction & Motivation Concrete Dropout (Extended Study) addresses a central problem in bayesiandl modeling with calibrated predictive…
Concrete Dropout (Extended Study) Introduction & Motivation Concrete Dropout (Extended Study) addresses a central problem in bayesiandl modeling with calibrated predictive…
concurrency control locking, concurrency control, control locking
Concurrency Control Locking Introduction & Motivation Concurrency Control Locking addresses a central problem in programming process control…
concurrent queue, concurrent hash map, fine grained locking, lock coupling, concurrent programming
Concurrent Data Structures is the design and implementation of data structures that support simultaneous access by multiple threads without data corruption, using fine-grained…
production
Condition-based maintenance is the maintenance policy that triggers service actions when measured equipment condition exceeds predefined thresholds - it replaces purely…
Conditional Average Treatment Effect CATE Modeling Introduction & Motivation Conditional Average Treatment Effect CATE Modeling addresses a central problem in causal inference…
neural architecture
Conditional Batch Normalization (CBN) is a batch normalization variant where the affine parameters (gamma, beta) are predicted by a conditioning input — allowing the normalization…
Conditional Image Colorization Introduction & Motivation Conditional Image Colorization addresses a central problem in generative generation and color-guidance-effectiveness…
model optimization
Conditional Computation is an approach that activates only selected model components for each input - It scales model capacity without proportional per-sample compute. What Is…
neural architecture
Conditional Computation is the neural network design paradigm where only a fraction of the model's total parameters are activated for any given input, fundamentally decoupling…
Conditional Computation and Mixture of Experts Routing Introduction & Motivation Conditional Computation and Mixture of Experts Routing addresses a central problem in reducing…
Conditional Computation and Mixture of Experts Routing (Extended Study) Introduction & Motivation Conditional Computation and Mixture of Experts Routing (Extended Study) addresses…
Conditional Computation and Mixture of Experts Routing (Extended Study) Introduction & Motivation Conditional Computation and Mixture of Experts Routing (Extended Study) addresses…
generative models
Conditional control inputs is the external signals that guide generation toward specified structure, geometry, or appearance constraints - they extend text prompting with explicit…
Conditional Diffusion Models Introduction & Motivation Conditional Diffusion Models addresses a central problem in generative modeling of diffusion data: how to Learn a generative…
Conditional Diffusion Models (Extended Study) Introduction & Motivation Conditional Diffusion Models (Extended Study) addresses a central problem in generative modeling…
Conditional Diffusion Models (Extended Study) Introduction & Motivation Conditional Diffusion Models (Extended Study) addresses a central problem in generative modeling…
domain adaptation
Conditional Domain Adaptation (CDAN) represents a massive, critical evolution over standard adversarial Domain Adaptation (like DANN) that actively prevents catastrophic "negative…
Conditional Domain Adversarial Networks CDAN Introduction & Motivation Conditional Domain Adversarial Networks CDAN addresses a central problem in transfer learning and (target)…
conditional execution and branch elimination, conditional execution branch, execution branch elimination, conditional execution, execution branch, branch elimination
Conditional Execution and Branch Elimination Introduction & Motivation Conditional Execution and Branch Elimination addresses a central problem in isa process control…
Conditional GAN Introduction & Motivation Conditional GAN addresses a central problem in generative modeling of diffusion data: how to Learn a generative model that produces…
Conditional GAN (Extended Study) Introduction & Motivation Conditional GAN (Extended Study) addresses a central problem in generative modeling of diffusion data: how to Learn…
Conditional GAN (Extended Study) Introduction & Motivation Conditional GAN (Extended Study) addresses a central problem in generative modeling of diffusion data: how to Learn…
Conditional Generation Methods Introduction & Motivation Conditional Generation Methods addresses a central problem in generative generation…
graph neural networks
Conditional Graph Gen is graph generation conditioned on target properties, context variables, or control tokens - It directs the generative process toward application-specific…
time series models
Conditional Independence is statistical criterion where variables become independent after conditioning on relevant factors. - It underpins causal graph discovery by identifying…