Glossary
171 abbreviations used across the taxonomy.
171 entries
| Abbreviation | Expansion |
|---|---|
| ActAdd | Activation Addition |
| AdaLoRA | Adaptive Low-Rank Adaptation |
| AEGIS | AI/ML Evaluation and Governance Infrastructure for Safety |
| AI | Artificial Intelligence |
| AL | Active Learning |
| API | Application Programming Interface |
| APO | Automatic Prompt Optimization |
| AWQ | Activation-aware Weight Quantization |
| BERT | Bidirectional Encoder Representations from Transformers |
| BitFit | BIas-Term FIne-Tuning |
| CAA | Contrastive Activation Addition |
| CAI | Constitutional AI |
| CAVIA | Context Adaptation VIA meta-learning |
| CDSS | Clinical Decision Support System |
| CE | Context Engineering |
| CI/CD | Continuous Integration/Continuous Deployment |
| CISPO | Clipped IS-weight Policy Optimization |
| CL | Continual Learning |
| CLIP | Contrastive Language-Image Pre-Training |
| CoT | Chain-of-Thought |
| CPO | Contrastive Preference Optimization |
| CPT | Continued Pre-Training |
| CPT | Context-Aware Prompt Tuning |
| CT | Continuous Training |
| DA | Domain Adaptation |
| DAPO | Decoupled clip and Dynamic sAmpling Policy Optimization |
| DARE | Drop And REscale |
| DELLA-Merging | Drop and rEscaLe via sampLing with mAgnitude |
| DIL | Domain-Incremental Learning |
| DL | Deep Learning |
| DoRA | Weight-Decomposed Low-Rank Adaptation |
| DP-FT | Differentially Private Fine-Tuning |
| DP-PEFT | Differentially Private Parameter-Efficient Fine-Tuning |
| DP-SGD | Differentially Private Stochastic Gradient Descent |
| DPO | Direct Preference Optimization |
| DPR | Dense Passage Retrieval |
| DSPy | Declarative Self-improving Python |
| DuDe | Dual Decomposition of Weights and Singular Value Low-Rank Adaptation |
| DyLoRA | Dynamic Low-Rank Adaptation |
| ECE | Expected Calibration Error |
| EDA | Easy Data Augmentation |
| EMR-Merging | Elect, Mask \& Rescale-Merging |
| EWC | Elastic Weight Consolidation |
| FDA | Food and Drug Administration |
| FedPara | Federated Parameterization |
| FILA | Fisher-Initialization of Low-rank Adapters |
| FL | Federated Learning |
| FLOP | Floating-Point Operation |
| FM | Foundation Model |
| FOMAML | First-Order Model Agnostic Meta Learning |
| FRIA | Fundamental Rights Impact Assessment |
| FSL | Few-Shot Learning |
| FT | Fine-Tuning |
| FTI | Feature-Training-Inference |
| FTL | Federated Transfer Learning |
| FV | Function Vector |
| GA | Gradient Ascent |
| GAN | Generative Adversarial Network |
| GD | Gradient Descent |
| GDPR | General Data Protection Regulation |
| GenFM | Generative Foundation Model |
| GMLP | Good Machine Learning Practice |
| GPAI | General-Purpose AI |
| GPT | Generative Pre-trained Transformer |
| GPTQ | Generative Pre-trained Transformers Quantization |
| GraphRAG | Graph Retrieval-Augmented Generation |
| GRPO | Group Relative Policy Optimization |
| GSPO | Group Sequence Policy Optimization |
| HippoRAG | Hippocampus-inspired Retrieval-Augmented Generation |
| IA$^3$ | Infused Adapter by Inhibiting and Amplifying Inner Activations |
| ICL | In-Context Learning |
| IMDRF | International Medical Device Regulators Forum |
| IPO | Identity Preference Optimization |
| ITI | Inference-Time Intervention |
| IVDR | In Vitro Diagnostic Regulation |
| KD | Knowledge Distillation |
| KL | Kullback-Leibler |
| KTO | Kahneman-Tversky Optimization |
| LATA | Layer-Aware Task Arithmetic |
| LEACE | LEAst-squares Concept Erasure |
| LEO | Latent Embedding Optimization |
| LLM | Large Language Model |
| LOKA | Large language mOdel Knowledge updAtes |
| LoKU | Low-rank Knowledge Unlearning |
| LoRA | Low-Rank Adaptation |
| LOSO | Leave One Seed Out |
| LP | Linear Probing |
| LwF | Learning without Forgetting |
| MAGPRUNE | Magnitude-based Pruning |
| MAKE | Memory-Associated Knowledge Editing |
| MallowsPO | Mallows-model Preference Optimization |
| MAML | Model-Agnostic Meta-Learning |
| MC | Monte Carlo |
| MCP | Model Context Protocol |
| MDR | Medical Device Regulation |
| MDSW | Medical Device Software |
| MEMIT | Mass-Editing Memory in a Transformer |
| MEND | Model Editor Networks with Gradient Decomposition |
| Meta-SGD | Meta Stochastic Gradient Descent |
| MiLoRA | Minor singular component based Low-Rank Adaptation |
| ML | Machine Learning |
| MLLM | Multimodal Large Language Model |
| MLOps | Machine Learning Operations |
| MLP | Multi-Layer Perceptron |
| MLR | Multivocal Literature Review |
| MMD | Maximum Mean Discrepancy |
| MoE | Mixture of Experts |
| MoELoRA | Low-Rank Adaptation as a Mixture of Experts |
| MTL | Multi-Task Learning |
| NAS | Neural Architecture Search |
| NIST | National Institute of Standards and Technology |
| NLP | Natural Language Processing |
| NMT | Neural Machine Translation |
| OPRO | Optimization by PROmpting |
| ORPO | Odds Ratio Preference Optimization |
| PCCP | Predetermined Change Control Plan |
| PE | Prompt Engineering |
| PEFT | Parameter-Efficient Fine-Tuning |
| PEPE | Periodic Extrapolation Positional Encodings |
| PI | Position Interpolation |
| PiSSA | Principal Singular values and Singular vectors Adaptation |
| PMET | Precise Model Editing in a Transformer |
| PMS | Post-Market Surveillance |
| PPO | Proximal Policy Optimization |
| PRF | Probabilistic Relevance Framework |
| PTQ | Post-Training Quantization |
| QAT | Quantization-Aware Training |
| QLoRA | Quantized Low-Rank Adaptation |
| RAFT | Retrieval-Augmented Fine-Tuning |
| RAFT | Reward rAnked FineTuning |
| RAG | Retrieval-Augmented Generation |
| ReFT | Representation Fine-Tuning |
| RepE | Representation Engineering |
| RepFM | Representation Foundation Model |
| RETRO | Retrieval-Enhanced Transformer |
| RL | Reinforcement Learning |
| RLAIF | Reinforcement Learning from AI Feedback |
| RLHF | Reinforcement Learning from Human Feedback |
| RLOO | REINFORCE Leave-One-Out |
| RLVR | Reinforcement Learning from Verifiable Rewards |
| RMF | Risk Management Framework |
| RMU | Representation Misdirection for Unlearning |
| ROME | Rank-One Model Editing |
| RoPE | Rotary Positional Embedding |
| rsLoRA | rank-stabilized Low-Rank Adaptation |
| RTN | Round-To-Nearest |
| SAE | Sparse AutoEncoder |
| SAPO | Step-Aligned Policy Optimization |
| Self-RAG | Self-Reflective Retrieval-Augmented Generation |
| Semi-SL | Semi-Supervised Learning |
| SERAC | Semi-Parametric Editing with a Retrieval-Augmented Counterfactual Model |
| SFT | Supervised Fine-Tuning |
| SGD | Stochastic Gradient Descent |
| SimPO | Simple Preference Optimization |
| SISA | Sharded, Isolated, Sliced, and Aggregated |
| SLERP | Spherical Linear Interpolation |
| SPIN | Self-Play fIne-tuNing |
| SSL | Self-Supervised Learning |
| STaR | Self-Taught Reasoner |
| SVD | Singular Value Decomposition |
| SWA | Stochastic Weight Averaging |
| TCAV | Testing with Concept Activation Vectors |
| TIES-MERGING | TrIm, Elect Sign \& Merge |
| TIL | Task-Incremental Learning |
| TL | Transfer Learning |
| TRL | Transformer Reinforcement Learning |
| TTA | Test-Time Adaptation |
| TTT | Test-Time Training |
| VeRA | Vector-based Random Matrix Adaptation |
| WMDP | Weapons of Mass Destruction Proxy |
| YaRN | Yet another RoPE extensioN |