29,259 technical terms and definitions
MIPS RISC, Stanford RISC, pipelined processor, workstation CPU, teaching architecture
MIPS is a RISC instruction-set architecture that originated from research at Stanford University in the early 1980s and became one of the most influential and widely studied RISC…
Maximal Interfered Retrieval MIR for Experience Replay Introduction & Motivation Maximal Interfered Retrieval MIR for Experience Replay addresses a central problem in continual…
optimization
Mirostat is an adaptive sampling algorithm that targets stable perplexity during generation - It is a core method in modern semiconductor AI serving and inference-optimization…
text generation
Mirostat sampling is the adaptive decoding algorithm that dynamically adjusts sampling behavior to target a desired output surprise or perplexity level - it provides…
mirror descent bregman divergence prediction, mirror descent bregman, descent bregman divergence, bregman divergence prediction, mirror descent, descent bregman
Mirror Descent Bregman Divergence Prediction with Machine Learning Introduction & Motivation Mirror Descent Bregman Divergence Prediction with Machine Learning addresses a central…
Mirror Descent Optimization Introduction & Motivation Mirror Descent Optimization addresses a central problem in optimization algorithm tuning and improvement: how to Optimize…
defects
Misfit Dislocations are linear crystal defects lying in the plane of a heteroepitaxial interface that partially relieve the biaxial strain produced by lattice mismatch between two…
neural architecture
Mish is a smooth, self-regularizing activation function defined as f(x) = x · tanh(softplus(x)) — combining the benefits of Swish-like self-gating with a bounded below property…
disinformation, information disorder, deepfakes, liar's dividend, Vosoughi Roy Aral 2018, AI-generated misinformation
Misinformation is false or misleading information that spreads regardless of whether the person sharing it knows it is wrong or intends harm. The word covers an honest mistake…
nlp
Misinformation detection is the AI/NLP task of identifying false or misleading information that is spread without deliberate intent to deceive. Unlike disinformation (which is…
mismatch threshold voltage prediction, mismatch threshold voltage, threshold voltage prediction, mismatch threshold, threshold voltage, voltage prediction
Mismatch Threshold Voltage Prediction with Machine Learning Introduction & Motivation Mismatch Threshold Voltage Prediction with Machine Learning addresses a central problem…
crystal misorientation, ebsd misorientation, disorientation analysis, kernel average misorientation, grain reference orientation deviation, misorientation distribution function
Two neighboring pixels can differ by half a degree because a crystal is bending, because the detector geometry drifts across the scan, or because orientation noise is being…
misr, advanced test & probe
MISR is a multiple-input signature register that compresses parallel test responses into compact signatures - Input responses are folded through feedback logic so large output…
MissForest Iterative Random Forest Imputation Introduction & Motivation MissForest Iterative Random Forest Imputation addresses a central problem in (subject) assurance…
Donald Rubin MCAR MAR MNAR, listwise deletion, imputation, multiple imputation, missing value indicator
Missing data is any value absent from a dataset where a value was expected — a survey question left blank, a sensor reading that failed to record, a database field that was never…
missing data imputation, missing data, data imputation
Missing Data Imputation Introduction & Motivation Missing Data Imputation addresses a central problem in timeseries process control and missing-value-imputation-method prediction…
Missing Data Imputation in Time Series Introduction & Motivation Missing Data Imputation in Time Series addresses a central problem in timeseries forecasting…
multimodal ai
Missing Modality Handling defines the critical suite of defensive architectural protocols engineered into Multimodal Artificial Intelligence to prevent immediate catastrophic…
Missing Value Detection and Pattern Analysis Introduction & Motivation Missing Value Detection and Pattern Analysis addresses a central problem in (subject) assurance and (target)…
impute, handle
Handling Missing Values is a critical data preprocessing step in machine learning because most algorithms cannot process NaN/Null values — requiring practitioners to choose…
Missingness Indicator Correlation and Pattern Heatmaps Introduction & Motivation Missingness Indicator Correlation and Pattern Heatmaps addresses a central problem in (subject)…
quality
Mistake-proofing is systematic implementation of controls that prevent, detect, or immediately signal process errors - Controls are embedded in workflow steps so deviations are…
quality & reliability
Mistake-Proofing is the design of processes and devices to prevent errors or detect them at the earliest possible moment - It is a core method in modern semiconductor quality…
foundation model
Mistral is an efficient open-source language model family featuring innovations like sliding window attention. Company: Mistral AI (French startup, founded by ex-DeepMind/Meta…
multi-die package, chiplet heterogeneous integration
Modern system-on-chip roadmaps increasingly abandon single monolithic silicon in favor of a mix and match chiplet strategy, where logic, memory, analog, and input/output functions…
dsp, digital signal processor, signal processing, mixed signal verification
Digital signal processing is the numerical manipulation of sampled signals to filter, transform, detect, compress, modulate, or reconstruct information. Every layer of an AI…
mixed clock domain synchronization complexity, mixed clock domain, clock domain synchronization, domain synchronization complexity, mixed clock, clock domain
Mixed Clock Domain Synchronization Complexity Introduction & Motivation Mixed Clock Domain Synchronization Complexity addresses a central problem in timing process control…
mixed-criticality task scheduling, mixed-criticality task, task scheduling
Mixed-Criticality Task Scheduling Introduction & Motivation Mixed-Criticality Task Scheduling addresses a central problem in real-time process control…
mixed criticality task scheduling, mixed criticality task, criticality task scheduling, mixed criticality, criticality task, task scheduling
Mixed Criticality Task Scheduling Introduction & Motivation Mixed Criticality Task Scheduling addresses a central problem in real-time-systems process control…
mixed-device scheduling
Mixed-Device Scheduling Introduction & Motivation Mixed-Device Scheduling addresses a central problem in ml-inference process control and scheduling-latency-milliseconds…
Branch-and-Cut for Mixed-Integer Linear Programming Introduction & Motivation Branch-and-Cut for Mixed-Integer Linear Programming addresses a central problem in optimization…
milp, ai safety
MILP (Mixed-Integer Linear Programming) Verification is the encoding of neural network verification problems as mixed-integer optimization problems — where ReLU activations are…
mixed integer programming (mip) prediction, mixed integer programming, integer programming (mip), programming (mip) prediction, mixed integer, integer programming, programming (mip)
Mixed Integer Programming (MIP) Prediction with Machine Learning Introduction & Motivation Mixed Integer Programming (MIP) Prediction with Machine Learning addresses a central…
manufacturing operations
Mixed Model Production is producing different product variants on the same line in an interleaved sequence - It supports demand variety without dedicated lines for each model.…
fp16, bf16, amp
Mixed-precision training is the standard recipe that lets modern models train in half the memory and roughly twice the throughput without losing accuracy. The idea is simple…
amp, automatic
Mixed-precision training is the standard recipe that lets modern models train in half the memory and roughly twice the throughput without losing accuracy. The idea is simple…
Automatic Mixed Precision AMP with Gradient Scaling Introduction & Motivation Automatic Mixed Precision AMP with Gradient Scaling addresses a central problem in optimization…
Mixed-Precision Quantization Introduction & Motivation Mixed-Precision Quantization addresses a central problem in reducing the size and cost of compression models: how to Reduce…
mixed-precision and quantization impact on hardware efficiency, mixed-precision quantization impact, quantization impact hardware, impact hardware efficiency, quantization impact
Mixed-Precision and Quantization Impact on Hardware Efficiency Introduction & Motivation Mixed-Precision and Quantization Impact on Hardware Efficiency addresses a central problem…
Mixed-Precision Quantization (Extended Study) Introduction & Motivation Mixed-Precision Quantization (Extended Study) addresses a central problem in reducing the size and cost…
Mixed-Precision Quantization (Extended Study) Introduction & Motivation Mixed-Precision Quantization (Extended Study) addresses a central problem in reducing the size and cost…
model optimization
Mixed-Precision Training is a training strategy that uses multiple numeric precisions to accelerate compute while preserving model quality - It lowers memory bandwidth…
mixed precision, precision training, bert fp16 fp32 nlp
Mixed Precision Training Introduction & Motivation Mixed precision: use FP16 for computation, FP32 for gradients. Faster training, lower memory. Applications: large model…
FP16 BF16 FP8, automatic mixed precision, gradient scaling, numerical stability
Mixed precision training combines lower-precision formats (FP16 or BF16) for the bulk of a network's compute-intensive operations with full FP32 precision retained for specific…
fp16 training, bfloat16 bf16, automatic mixed precision amp, loss scaling gradient
Mixed Precision Training is the technique of using lower-precision floating-point formats (FP16 or BF16) for most computations while maintaining FP32 precision for critical…
fp16 training, bfloat16 training, automatic mixed precision amp, loss scaling
Mixed Precision Training is the technique that uses lower precision (FP16 or BF16) for most computations while maintaining FP32 for critical operations — reducing memory usage…
automatic mixed precision amp, loss scaling fp16 training, half precision training optimization, mixed precision gradient underflow
Mixed Precision Training is the optimization technique that uses lower-precision floating-point formats (FP16 or BF16) for the majority of training computations while maintaining…
mixed-precision training methods, mixed-precision training, training methods
Mixed-Precision Training Methods Introduction & Motivation Mixed-Precision Training Methods addresses a central problem in ml-training process control…
mixed precision training, mixed precision, precision training
Mixed Precision Training Introduction & Motivation Mixed Precision Training addresses a central problem in systems process control and precision-loss-accuracy-degradation…
mixed-precision training numerical stability and convergence, mixed-precision training numerical, training numerical stability, numerical stability convergence, training numerical
Mixed-Precision Training Numerical Stability and Convergence Introduction & Motivation Mixed-Precision Training Numerical Stability and Convergence addresses a central problem…