← Back to AI Factory Chat

AI Factory Glossary

3,145 technical terms and definitions

A B C D E F G H I J K L M N O P Q R S T U V W X Y Z All
Showing page 51 of 63 (3,145 entries)

scope 1 emissions, environmental & sustainability

Scope 1 emissions are direct greenhouse gas emissions from fab operations including process gases combustion and fugitive emissions.

scope 2 emissions, environmental & sustainability

Scope 2 emissions are indirect emissions from purchased electricity steam heating and cooling consumed by fab operations.

scope 3 emissions, environmental & sustainability

Scope 3 emissions include indirect supply chain emissions from materials transportation business travel and product use.

score distillation, multimodal ai

Score distillation sampling uses diffusion model gradients to optimize 3D representations.

score matching for ebms, generative models

Train by matching score functions.

score matching,generative models

Train by matching data scores.

score-based generative models via sdes, generative models

Unified framework connecting scores and diffusion.

score-based generative models,generative models

Generate by learning score function.

score-cam, explainable ai

Generate CAMs without gradients.

scoring functions, healthcare ai

Evaluate quality of docking poses.

scribble conditioning, multimodal ai

Scribble conditioning interprets rough sketches as spatial guidance for generation.

scribble control, generative models

Condition on rough sketches.

scrubber system, environmental & sustainability

Scrubber systems remove hazardous gases from process tool exhaust through chemical reaction or adsorption before atmospheric release.

sd upscale, sd, generative models

Stable Diffusion upscaling.

sdr, sdr, failure analysis advanced

Soft Defect Localization uses stimulation and imaging techniques to reveal intermittent or parametric failures.

se transformer, se(3), graph neural networks

SE(3) Transformer processes 3D point clouds and molecular structures with equivariance to rotation and translation through spherical harmonics.

se-transformers, scientific ml

Transformers with 3D symmetries.

se3-equivariant gnn, graph neural networks

SE(3)-equivariant GNNs process molecular geometries maintaining equivariance to rotations translations and reflections.

seamless tiling, generative models

Generate seamlessly tileable images.

search space design, neural architecture search

Search space design determines the set of possible architectures balancing expressiveness with search efficiency.

seasonal state space, time series models

Seasonal state space models represent seasonality through trigonometric or dummy variable state equations.

secure aggregation, training techniques

Secure aggregation computes collective statistics without revealing individual contributions.

secure multi-party, training techniques

Secure multi-party computation allows joint computation without revealing private inputs.

seebeck effect fa, failure analysis advanced

Seebeck effect imaging detects voltage gradients from thermal variations revealing resistive defects and current paths.

seeds yield model, yield enhancement

Seeds yield model accounts for systematic and random defects with separate density and clustering parameters.

segment-level recurrence, llm architecture

Process segments recurrently.

segmentation control, generative models

Condition on semantic maps.

selective knowledge distillation, model compression

Choose which knowledge to transfer.

selective prediction, ai safety

Selective prediction abstains from answering when confidence is insufficient.

selective prediction,ai safety

Choose when to abstain from prediction.

selective ssm, llm architecture

Selective SSMs dynamically adjust state transitions based on input content.

self-alignment, training techniques

Self-alignment improves models using their own generated critiques and revisions.

self-attention in capsules,neural architecture

Combine capsules with attention.

self-attentive hawkes, time series models

Self-attentive Hawkes processes combine transformer architectures with temporal point process likelihoods for scalable event sequence modeling.

self-critiquing, ai safety

Model critiques its own outputs.

self-distillation, model compression

Distill model into itself.

self-distillation, model optimization

Self-distillation uses model's own predictions as soft targets improving generalization.

self-ensembling for domain adaptation, domain adaptation

Ensemble predictions for consistency.

self-explaining neural networks, senn, explainable ai

Architectures designed for interpretability.

self-gating, neural architecture

Use gates controlled by input.

self-heating modeling, simulation

Model temperature rise in device.

self-instruct, training techniques

Self-Instruct generates training data by prompting models to create instructions and responses.

self-monitoring, ai agents

Self-monitoring tracks progress evaluating whether actions advance toward goals.

self-paced learning, advanced training

Self-paced learning automatically selects training samples based on current model capability starting with easier examples.

self-paced learning, machine learning

Let model select training order.

self-supervised gnn, graph neural networks

Self-supervised graph neural networks learn representations through pretext tasks without labeled data.

self-supervised pre-training for vit, computer vision

DINO MAE BEiT methods.

self-training, advanced training

Self-training is a semi-supervised method where models generate pseudo-labels for unlabeled data to augment training sets iteratively.

self-training,semi-supervised learning

Use model predictions on unlabeled data as training labels.

selfies, selfies, chemistry ai

Alternative molecular representation.