29,258 technical terms and definitions
damascene, dual-damascene, dual damascene, single damascene, copper damascene, damascene inlay, damascene integration, damascene interconnect, damascene metallization, copper inlay process, via-first, trench-first, semi-damascene
The damascene process exists because copper cannot be plasma-etched. Aluminium forms volatile AlCl₃ at room temperature in a chlorine plasma, so for decades the interconnect…
dual damascene, copper damascene, inlaid metallization, copper interconnect
The damascene process, whether single or dual damascene, defines metal interconnect features by first etching a pattern into a dielectric layer and then filling that pattern…
interconnect controlling line width in damascene interconnect processes, interconnect controlling line, controlling line width, line width damascene, width damascene interconnect
Interconnect: Controlling line width in damascene interconnect processes Introduction & Motivation Interconnect: Controlling line width in damascene interconnect processes…
damascene process line width control, damascene process line, process line width, line width control, damascene process, process line, line width, width control
Damascene Process Line Width Control with Machine Learning Introduction & Motivation Damascene Process Line Width Control with Machine Learning addresses a central problem…
dan, do anything now, ai safety
DAN (Do Anything Now) is the most widely known jailbreak prompt framework that attempts to make ChatGPT bypass its safety restrictions by role-playing as an unrestricted AI…
jailbreak, llm safety, adversarial prompts, prompt injection, ai safety, alignment, ai security
DAN prompts are jailbreaking techniques that attempt to bypass AI safety guardrails by instructing the model to role-play as "Do Anything Now" — adversarial prompts that frame…
dann, domain adaptation
DANN (Domain-Adversarial Neural Network) is the seminal, groundbreaking architecture defining modern Deep Domain Adaptation, mathematically forcing a feature extractor to learn…
dare, model merging
DARE (Drop and Rescale) is a model merging technique that randomly drops (zeros out) a fraction of fine-tuned parameter changes and rescales the remaining ones — reducing…
dark current noise, dark current nonuniformity
Dark current is the current a photodiode or image sensor pixel produces with absolutely no light reaching it — thermally generated electron-hole pairs crossing the same depletion…
dark field bright field inspection correlation prediction with machine learning, dark field bright field inspection correlation prediction machine learning
Dark-Field/Bright-Field Inspection Correlation Prediction with Machine Learning Introduction & Motivation Dark-Field/Bright-Field Inspection Correlation Prediction with Machine…
dark field inspection sensitivity drift prediction machine learning, dark field, inspection sensitivity drift, dark field inspection sensitivity drift prediction
Dark Field Inspection Sensitivity Drift Prediction with Machine Learning Introduction & Motivation Dark Field Inspection Sensitivity Drift Prediction with Machine Learning…
model compression
Dark Knowledge is the rich information contained in a teacher model's soft output distribution — the relative probabilities assigned to incorrect classes reveal the model's…
model optimization
Dark Knowledge is informative class-probability structure in teacher outputs that reveals inter-class relationships - It captures nuanced uncertainty patterns not present in hard…
power wall, utilization wall, power density, dennard scaling end, thermal ceiling, utilization ceiling, dark silicon problem
Dark silicon is the fraction of a chip that must be left unpowered — kept dark — at any given moment because the whole die cannot be switched on within its power and thermal…
dark silicon active core percentage, dark silicon active, silicon active core, active core percentage, dark silicon, silicon active, active core, core percentage
Dark Silicon Active Core Percentage Introduction & Motivation Dark Silicon Active Core Percentage addresses a central problem in design process control…
dark silicon and active core percentage, dark silicon active, silicon active core, active core percentage, dark silicon, silicon active, active core, core percentage
Dark Silicon and Active Core Percentage Introduction & Motivation Dark Silicon and Active Core Percentage addresses a central problem in design process control…
dark silicon percentage, dark silicon, silicon percentage
Dark Silicon Percentage Introduction & Motivation Dark Silicon Percentage addresses a central problem in design process control and active-core prediction: how to Predict…
dark silicon management, dark silicon, silicon management
Dark Silicon Management Introduction & Motivation Dark Silicon Management addresses a central problem in design process control and core-pct prediction: how to Predict core-pct…
dark silicon and active core percentage, dark silicon active, silicon active core, active core percentage, dark silicon, silicon active, active core, core percentage
Dark Silicon and Active Core Percentage Introduction & Motivation Dark Silicon and Active Core Percentage addresses a central problem in design process control…
metrology
Darkfield Inspection is a semiconductor metrology technique that illuminates wafers at oblique angles and collects only scattered light from defects — blocking the specular…
Dartmouth Summer Research Project on Artificial Intelligence, Dartmouth workshop, Hanover New Hampshire 1956, Logic Theorist demonstration, birth of AI, McCarthy Minsky Rochester Shannon, Newell Simon Shaw
The Dartmouth conference of 1956 was a summer workshop on artificial intelligence held at Dartmouth College in Hanover, New Hampshire, organized by John McCarthy, Marvin Minsky…
Dartmouth Summer Research Project proposal, August 31 1955, Rockefeller Foundation grant request, McCarthy Minsky Rochester Shannon, simulate intelligence conjecture, seven research topics
The Dartmouth proposal is the document, dated August 31, 1955, in which John McCarthy, Marvin Minsky, Nathaniel Rochester and Claude Shannon asked the Rockefeller Foundation…
darts, neural architecture search
DARTS is a differentiable neural-architecture-search method that relaxes discrete architecture choices into continuous optimization - Architecture parameters and network weights…
DARTS Differentiable Architecture Search Introduction & Motivation DARTS Differentiable Architecture Search addresses a central problem in automated design of nas architectures…
DARTS Differentiable Architecture Search (Extended Study) Introduction & Motivation DARTS Differentiable Architecture Search (Extended Study) addresses a central problem…
DARTS Differentiable Architecture Search (Extended Study) Introduction & Motivation DARTS Differentiable Architecture Search (Extended Study) addresses a central problem…
DARTS Differentiable Architecture Search (Extended Study) Introduction & Motivation DARTS Differentiable Architecture Search (Extended Study) addresses a central problem…
necking crystal growth, dislocation free silicon, czochralski neck, seed necking, william dash crystal
Dash necking is the deliberate step of pulling a growing silicon crystal down to a thin thread just a few millimeters wide right after the seed touches the melt, so that…
parallel, distributed
Dask is the parallel computing library for Python that scales NumPy, Pandas, and Scikit-Learn workflows from a single workstation to a cluster by chunking data into manageable…
Python, parallel, computing, distributed, task, scheduler, lazy
Dask Python Parallel Computing is a flexible Python library providing parallel computing via task graphs and lazy evaluation, enabling scalable data processing on single machines…
dask array dataframe, dask scheduler, dask delayed computation, dask distributed cluster, dask vs spark, dask task graph, dask ml, dask chunk size, dask out of core
Dask scales Python's data stack—NumPy, Pandas, scikit-learn—beyond single-machine memory by representing a computation as a directed acyclic task graph that a scheduler executes…
parallelism, all-reduce, optimization, algorithms
Data is the raw material every machine learning system is built from — not yet a feature, not yet a label, just the unprocessed record of something that happened: a pixel value…
exploratory data analysis, confirmatory data analysis, John Tukey, quantitative vs qualitative analysis, confirmation bias data
Data analysis is the process of inspecting, cleaning, and modeling data to surface conclusions that support a decision, and it sits one step narrower than data science: where data…
machine intelligence, ai, ai systems, data analytics, machine learning
artificial intelligence is the broad field of building machines that perceive, reason, learn, generate, plan, and act toward goals. AI spans symbolic systems, machine learning…
data
Data labeling turns raw examples into the supervised signal a model trains on, and the single hardest thing to internalize about it is that a label is a measurement, not ground…
privacy
Data anonymization is the process of removing or modifying personally identifiable information (PII) from datasets so that individuals cannot be identified from the remaining…
training techniques
Data Anonymization is process that irreversibly removes identifying information so individuals cannot be reasonably reidentified - It is a core method in modern semiconductor AI…
data anonymization techniques, data anonymization, anonymization techniques
Data Anonymization Techniques Introduction & Motivation Data Anonymization Techniques addresses a central problem in ml-ops process control and anonymization-time-per-record…
model training
Data augmentation creates new training examples from existing ones through label-preserving transformations — a flipped cat is still a cat, a back-translated sentence still means…
image augmentation, augmentation techniques
Data Augmentation — artificially expanding the training dataset by applying random transformations, improving generalization without collecting more data. Common Techniques…
training data expansion, mixup, cutmix, randaugment, synthetic data
Data augmentation creates additional training variation by transforming, mixing, simulating, or generating examples while preserving or deliberately modifying target semantics.…
augmentation strategy training, mixup cutmix augmentation, autoaugment randaugment, synthetic data augmentation
Data Augmentation is the training regularization technique that artificially expands the effective size and diversity of a training dataset by applying label-preserving…
augmentation strategy, mixup cutmix, augmentation pipeline, randaugment
Data Augmentation is the training-time technique that artificially expands the effective dataset size by applying random transformations to training examples — creating modified…
augmentation strategy training, cutout mixup cutmix, autoaugment randaugment, augmentation generalization overfitting
Data Augmentation in Deep Learning is the training regularization technique that artificially expands the effective training dataset by applying random transformations to input…
data augmentation generation, data augmentation, augmentation generation
Data Augmentation Generation Introduction & Motivation Data Augmentation Generation addresses a central problem in nlp process control and data-augmentation-text-generation…
randaugment augmentation policy, augmax robust augmentation, data augmentation deep learning, augmentation strategy training
Data Augmentation Strategies (Mixup, CutMix, RandAugment, AugMax) is the practice of applying transformations to training data to artificially increase dataset diversity…
data augmentation, mixup cutmix augmentation, cifar passed, data augmentation mixup cutmix & augmentation strategies, data augmentation mixup, augmentation mixup cutmix, mixup cutmix
Data Augmentation: Mixup, CutMix & Augmentation Strategies Introduction & Motivation Data Augmentation: increase training data variety. Mixup: linear interpolation in input space.…
data augmentation, mixup cutmix consistency, 1m 4d 50k cifar kl nlp, data augmentation mixup cutmix & consistency regularization, data augmentation mixup, augmentation mixup cutmix
Data Augmentation: Mixup, CutMix & Consistency Regularization Introduction & Motivation Data augmentation artificially increases training data via transformations. Mixup…
training techniques
Data Augmentation Privacy is augmentation strategy that improves model robustness while minimizing disclosure of identifiable training information - It is a core method in modern…
data augmentation, augmentation strategies, cifar nlp sst
Data Augmentation Strategies Introduction & Motivation Data Augmentation: expand training data synthetically. Image augmentation, text augmentation. Applications: improve…