AI Factory Glossary

ppap, ppap, quality & reliability

Production Part Approval Process validates supplier capability to meet requirements.

ppk index, quality & reliability

Ppk index combines overall performance with centering for long-term capability.

ppm targets, ppm, quality

Quality goals in parts per million.

ppo with adaptive kl, ppo, reinforcement learning

Adapt KL penalty.

ppo with clipping, ppo, reinforcement learning

Standard PPO variant.

ppo,policy gradient,actor critic

PPO is a stable policy gradient algorithm for RL. Used in RLHF to fine-tune LLMs from human preferences.

ppo,policy gradient,algorithm

PPO (Proximal Policy Optimization) is stable RL algorithm. Clipped objective. Used in RLHF.

pq,product quantization,compress

Product quantization compresses vectors. Trade accuracy for memory.

pr description,pull request,summarize

Summarize pull requests automatically. Explain changes.

pragmatics in nlp, nlp

Understand context-dependent meaning.

pre commit,hooks,quality

Pre-commit runs hooks before commit. Linting, formatting.

pre-aligner, manufacturing operations

Pre-aligners roughly position wafers before precision alignment.

pre-amorphization, process integration

Pre-amorphization implants destroy crystal structure before dopant implantation enabling uniform amorphous layers.

pre-conditioning test, reliability

Simulate real-world before reliability.

pre-control charts, spc

Simple alternative to SPC.

pre-emphasis, signal & power integrity

Pre-emphasis boosts high-frequency content at transmitter compensating for channel attenuation.

pre-layernorm vs post-layernorm, transformer

Placement of normalization in transformer.

pre-norm benefits, transformer

Why pre-norm helps transformer training.

pre-tokenization, nlp

Split before subword tokenization.

pre-training data scale for vit, computer vision

Billions of images for large ViTs.

precession electron diffraction, ped, metrology

Reduce dynamical effects in diffraction.

precious metal recovery, environmental & sustainability

Precious metal recovery reclaims gold silver and platinum from electronic waste.

precipitate growth, process

Enlargement of precipitates.

precipitation kinetics, process

Time evolution of precipitation.

precision at k, evaluation

Fraction of top K that are relevant.

precision at k,evaluation

Fraction of top-K that are relevant.

precision medicine,healthcare ai

Tailor treatment to individual patients.

precision-recall tradeoff in moderation, ai safety

Balance between catching all vs accuracy.

precision,metrology

Repeatability of measurements.

precondition inference,software engineering

Determine required conditions for functions.

precursor detection, reliability

Detect impending failures.

precursor,cvd

Chemical vapor that reacts on surface to form film.

predictability of emergence, theory

Whether emergent abilities are predictable.

prediction interval, quality & reliability

Prediction intervals provide ranges likely containing future observations.

prediction intervals,statistics

Range containing true value with given probability.

prediction set,statistics

Set of possible predictions with coverage guarantee.

predictive maintenance, manufacturing operations

Predictive maintenance uses condition monitoring to anticipate failures enabling proactive intervention.

predictive maintenance, production

Maintenance based on condition monitoring.

predictive maintenance,production

Use data analytics to predict failures before they occur.

predictive metrology, metrology

Forecast future metrology results.

preemptible instance training, infrastructure

Use interruptible cloud instances.

prefect,workflow,modern

Prefect is modern workflow orchestration. Simpler than Airflow. Python-native.

preference dataset, training techniques

Preference datasets contain ranked or compared responses for reward model training.

preference learning, training techniques

Preference learning trains models from human comparisons of response quality.

preference learning,rlhf

Learn what humans prefer from comparisons.

preference-based rl, rlhf

Learn from pairwise preferences.

prefix caching, llm optimization

Prefix caching stores common prompt beginnings for reuse across requests.

prefix caching,prompt cache,reuse

Prefix caching stores KV cache for common prefixes. Reuse for same system prompt. Reduces first-token latency.

prefix language modeling, foundation model

Bidirectional encoding with autoregressive decoding.

prefix sharing, inference

Share common prompt prefixes.