Alphabetical Table of Contents

Concepts A–Z. † marks a provisional concept identity; ◆ marks a concept new for the 2025 Collection (origin N). Part numbers refer to the thematic parts.

Concept Part Records
Advanced Reasoning Models (Cognitive Hybrid Architectures) † I AIK-2025-S-P-ADVREASON
Agentic AI V AIK-2025-E-P-AGENTICAI, AIK-2025-E-P-AGENTICAI-02
AgentScen — LLM-Based External World Models V AIK-2025-S-P-AGENTSCEN
AI Alignment IV AIK-2025-S-P-ALIGNMENT
AI Hallucination IV AIK-2025-S-P-HALLUC
AI Washing VI AIK-2025-S-P-AIWASHING
Alignment Faking (Deceptive Alignment) IV AIK-2025-E-P-ALIGNFAKE
Alignment Patterns: Constitutional, Preference-Modeling, & Red-Teaming IV AIK-2025-S-P-ALIGNPATTERN
Arithmetic and Reasoning Heads † IV AIK-2025-S-P-ARITHHEADS
Artificial Hivemind VI AIK-2025-E-P-HIVEMIND
Bidirectional Human-AI Alignment IV AIK-2025-S-P-BIDIRALIGN
Chain-of-Thought Monitorability ◆ IV AIK-2025-F-N-COTMONITOR
Chain-of-Thought Prompting III AIK-2025-F-P-COTPROMPT
Chain-of-Verification (CoVe) III AIK-2025-S-P-COVE
Circuit Sparsity IV AIK-2025-F-P-CIRCUITSP
Compound AI Systems V AIK-2025-E-P-COMPOUNDAI, AIK-2025-S-P-COMPOSABLE
Constitutional AI IV AIK-2025-F-P-CONSTAI, AIK-2025-S-P-CONSTAI
Constitutional AI 2.0 † IV AIK-2025-S-P-CAI2
Context Engineering ◆ V AIK-2025-F-N-CONTEXTENG
Context Rot ◆ V AIK-2025-F-N-CONTEXTROT
Continuous Autoregressive Language Models (CALM) I AIK-2025-F-P-CALM
Controllable Generation Frameworks III AIK-2025-S-P-CTRLGEN
Delta Learning Hypothesis II AIK-2025-S-P-DELTALEARN
Distillation at Scale / Student-Teacher Cascades II AIK-2025-S-P-DISTILSCALE
Drop-Upcycling (Sparse MoE Training) II AIK-2025-S-P-DROPUPCYCLE
Dynamic Planning for LLM Agents V AIK-2025-S-P-DYNPLAN
Dynamical Regularization (in Diffusion) II AIK-2025-S-P-DYNREG
E-Processes / Game-Theoretic Statistics II AIK-2025-E-P-EPROCESS
Edge AI VI AIK-2025-S-P-EDGEAI, AIK-2025-S-P-EDGEAI-02
Embodied AI (as a Paradigm) VI AIK-2025-S-P-EMBODIEDAI
Embodied Foundation Models VI AIK-2025-S-P-EMBODIEDFM
Emergent Misalignment ◆ IV AIK-2025-F-N-EMERGENTMIS
EvoAgent — Continual World-Model Agents V AIK-2025-S-P-EVOAGENT
Explainable AI (XAI) IV AIK-2025-S-P-XAI
Federated & On-Device Model Training II AIK-2025-S-P-FEDONDEVICE
Frugal AI VI AIK-2025-S-P-FRUGALAI, AIK-2025-S-P-FRUGALAI-02
Gated Attention I AIK-2025-E-P-GATEDATTN
Generative Engine Optimisation (GEO) VI AIK-2025-S-P-GEO
Generative Feedback Loop (Self-Adaptation) II AIK-2025-S-P-GENFEEDBACK
Generative User Interfaces (GenUI) V AIK-2025-F-P-GENUI
GraphRAG V AIK-2025-S-P-GRAPHRAG
High-Dimensional Learning Dynamics II AIK-2025-S-P-HDLD
Hyperautomation VI AIK-2025-S-P-HYPERAUTO
KV-Cache Sharing (On-Device Inference) III AIK-2025-E-P-KVSHARE
Large Reasoning Models (LRMs) I AIK-2025-S-P-LRM, AIK-2025-E-P-LRM
Large-Scale Population Agent Simulation (LLM Archetypes) V AIK-2025-S-P-POPAGENTSIM
LExI — Layer-Adaptive Expert Inference I AIK-2025-S-P-LEXI
Long-Context & Windowing Techniques I AIK-2025-S-P-LONGCONTEXT
Long-Horizon LLM Planning & Efficiency Metrics † V AIK-2025-S-P-LHPLANMETRIC
Mechanistic Interpretability Tools IV AIK-2025-E-P-MECHINTERP
Memorization Sinks II AIK-2025-S-P-MEMSINK
Mixture of Depths I AIK-2025-F-P-MOD
Mixture of Experts (MoE) I AIK-2025-F-P-MOE, AIK-2025-S-P-MOE, AIK-2025-S-P-MOE-02
Mixture of Latent Experts (MoLE / MoLAE) I AIK-2025-S-P-MOLE
Model Context Protocol (MCP) V AIK-2025-S-P-MCP, AIK-2025-E-P-MCP
Model Safety Evaluation Suites & LLM-Based Evaluators IV AIK-2025-S-P-SAFETYEVAL
Multi-Agent Debate Systems III AIK-2025-S-P-MADEBATE
Multimodal Native Architecture I AIK-2025-S-P-MMNATIVE
Nested Learning (Continual AI) II AIK-2025-S-P-NESTEDLEARN
On-Policy Distillation II AIK-2025-F-P-ONPOLDISTIL
Parameter-Efficient Fine-Tuning (PEFT) evolution II AIK-2025-S-P-PEFT
Plan-and-Act Agent Framework V AIK-2025-S-P-PLANACT
Prefill–Decode Disaggregation III AIK-2025-F-P-PDDISAGG
Programmatic Representations (for Agents) V AIK-2025-S-P-PROGREP
Prompt Injection Defense Layers IV AIK-2025-S-P-PIDEFENSE
Quantization & Ultra-Low-Precision Inference III AIK-2025-S-P-QUANTINFER
Retrieval-Augmented Generation (RAG) V AIK-2025-S-P-RAG, AIK-2025-S-P-RAG-02
RLAIF — Reinforcement Learning from AI Feedback II AIK-2025-S-P-RLAIF, AIK-2025-S-P-RLAIF-02
RLVR (Reinforcement Learning with Verifiable Rewards) ◆ II AIK-2025-F-N-RLVR
Scalable Oversight & Synthetic Critiques IV AIK-2025-E-P-SCALEOVER, AIK-2025-S-P-SCALEOVER
Scaling Laws for Sparse Models & Efficiency Metrics I AIK-2025-S-P-SPARSESCALE
Scheming (AI) ◆ IV AIK-2025-F-N-SCHEMING
Self-Supervised RL II AIK-2025-S-P-SSRL
Shadow AI VI AIK-2025-S-P-SHADOWAI
Shared Autonomy (Robotics) VI AIK-2025-S-P-SHAREDAUTO
Small Language Models (SLMs) I AIK-2025-S-P-SLM
Social Choice Fairness VI AIK-2025-S-P-SOCCHOICE
Soft Robotics / Dexterous Manipulation VI AIK-2025-S-P-SOFTROBOT
Speculative Decoding III AIK-2025-S-P-SPECDECODE
Synthetic Data II AIK-2025-S-P-SYNDATA
Synthetic Data Quality Filtering II AIK-2025-S-P-SYNDATAFILT
Test-Time Compute Scaling III AIK-2025-E-P-TTCSCALE
Theoretical Learning Dynamics of MoE † I AIK-2025-S-P-MOETHEORY
Tokenization-Centric Design I AIK-2025-S-P-TOKDESIGN
Vector Databases V AIK-2025-S-P-VECTORDB
Vertical AI Integration VI AIK-2025-S-P-VERTICALAI
Vibe Coding V AIK-2025-S-P-VIBECODE
World Model (LLM-Integrated) I AIK-2025-E-P-WORLDMODEL, AIK-2025-S-P-WORLDMODEL