Skip to content

Call Me Almanac

Home Library Topics Learn Editions Ask

Explore the corpus

Search Entities Events

Get the almanac in your inbox

Call Me Almanac·An AI-research almanac

About Library Editions Newsletter Ask Contact Privacy Terms RSS

© 2026 Call Me Almanac · Content is AI-generated and may be inaccurate.

Library

Plain-language guides to the models, labs, and ideas shaping AI — synthesized from the corpus.

All Topic guides Concept guides Entity guides

LoRAConcept

LoRA: How to Teach a Giant AI New Tricks Without Rebuilding It

Read asBeginner In-depth

ReActConcept

ReAct: How AI Agents Think and Act at the Same Time

Read asBeginner In-depth

GRPOConcept

GRPO: The Reinforcement Learning Trick Making AI Smarter at Reasoning

Read asBeginner In-depth

AdamWConcept

AdamW: The Default Optimizer Powering Modern AI Training

Read asBeginner In-depth

Preparedness FrameworkConcept

OpenAI's Preparedness Framework: A Plain-Language Guide to AI Safety Guardrails

Read asBeginner In-depth

MCPConcept

MCP (Model Context Protocol): The Universal Plug Standard for AI Agents

Read asBeginner In-depth

Proximal Policy OptimizationConcept

Proximal Policy Optimization (PPO): The Algorithm That Trains AI to Learn from Feedback

Read asBeginner In-depth

GRPO (Group Relative Policy Optimization)Concept

GRPO: The Reinforcement Learning Trick Behind Smarter AI Reasoning

Read asBeginner In-depth

Differential PrivacyConcept

Differential Privacy: A Mathematical Promise That Your Data Won't Be Exposed

Read asBeginner In-depth

Reinforcement LearningConcept

Reinforcement Learning: How AI Learns by Doing

Read asBeginner In-depth

Mixture of ExpertsConcept

Mixture of Experts: How AI Models Do More by Using Less

Read asBeginner In-depth

Constitutional AIConcept

Constitutional AI: Teaching Models to Follow Principles, Not Just Rules

Read asBeginner In-depth

Chain-of-Thought ReasoningConcept

Chain-of-Thought Reasoning: Teaching AI to Show Its Work

Read asBeginner In-depth

Diffusion ModelsConcept

Diffusion Models: How AI Learns to Paint by Unpainting

Read asBeginner In-depth

Direct Preference Optimization (DPO)Concept

Direct Preference Optimization (DPO): Aligning AI Without a Reward Model

Read asBeginner In-depth

knowledge distillationConcept

Knowledge Distillation: Teaching Small Models to Punch Above Their Weight

Read asBeginner In-depth

LLM-as-a-JudgeConcept

LLM-as-a-Judge: Using AI to Grade AI

Read asBeginner In-depth

mechanistic interpretabilityConcept

Mechanistic Interpretability: Looking Inside the AI Black Box

Read asBeginner In-depth

MambaConcept

Mamba: The Attention-Free Architecture That Scales Without Slowing Down

Read asBeginner In-depth

PPOConcept

PPO: The Reinforcement Learning Algorithm That Taught AI to Learn from Feedback

Read asBeginner In-depth

prompt injectionConcept

Prompt Injection: The Security Threat Hiding in Plain Text

Read asBeginner In-depth

Model Context ProtocolConcept

Model Context Protocol (MCP): The Universal Plug for AI Agents

Read asBeginner In-depth

Reinforcement Learning from Human FeedbackConcept

Reinforcement Learning from Human Feedback (RLHF): Teaching AI to Do What You Mean

Read asBeginner In-depth

Reinforcement Learning with Verifiable RewardsConcept

Reinforcement Learning with Verifiable Rewards (RLVR): Teaching AI to Check Its Own Work

Read asBeginner In-depth

Retrieval-Augmented GenerationConcept

Retrieval-Augmented Generation (RAG): Giving AI a Library Card

Read asBeginner In-depth

scalable oversightConcept

Scalable Oversight: Teaching AI to Help Humans Stay in Charge

Read asBeginner In-depth

supervised fine-tuningConcept

Supervised Fine-Tuning: Teaching an AI to Do Your Job

Read asBeginner In-depth

speculative decodingConcept

Speculative Decoding: Making AI Faster Without Changing the Answer

Read asBeginner In-depth

Vision-Language ModelsConcept

Vision-Language Models: Teaching AI to See and Read at Once

Read asBeginner In-depth