Reasoning model
A reasoning model is an AI model designed or optimized to perform multistep inference or analysis before producing an answer or result.
Reinforcement Learning from Human Feedback
A reasoning language model is a language model optimized to solve problems through explicit or implicit multi-step reasoning before producing a final answer.
- How it is used
- Improves math, code, science, and other verifiable-task performance through RLVR and added generation-time computation.
- How deep it goes
- Explains the RLVR loop, verifier rewards, training practices, historical methods, and evaluation changes without detailing a full implementation.
- Examples
- Contrasts an opportunity-cost explanation with checkable math or code answers, and uses formatted multiple-choice chain-of-thought prompting.
- What you can do after
- Evaluate/understand how verifier-based reinforcement learning and inference-time scaling produce reasoning-oriented model behavior.
Read how Reinforcement Learning from Human Feedback explains reasoning model →
Build a Reasoning Model (From Scratch)
A language model optimized to deliberate over complex problems, often producing intermediate steps or using tools to improve answer quality.
- How it is used
- Deploys intermediate-token generation for complex mathematics, coding, logic, planning, and tool-selection tasks.
- How deep it goes
- Defines visible and hidden intermediate steps, contrasts pattern matching with explicit logic, and outlines inference and training methods without implementing them.
- Examples
- Uses contradictory bird premises, chain-of-thought diagrams, token-by-token generation, roadmap figures, and math-verifier workflows.
- What you can do after
- Evaluate/understand when intermediate reasoning improves accuracy and how verification can measure or train reasoning behavior.
Read how Build a Reasoning Model (From Scratch) explains reasoning model →
AI Agents in Action, Second Edition
A reasoning model is a language model that explicitly allocates computation to multistep deliberation before producing its final output.
- How it is used
- Supports agent planning and other difficult multistep steps while avoiding unnecessary deliberation during routine tool-constrained work.
- How deep it goes
- Explains the deliberation-versus-latency trade-off and configurable effort, without detailing model internals.
- Examples
- Uses a bounded planning-agent listing with a checklist, tool observations, structured status, and a maximum-turn setting.
- What you can do after
- Operate/configure it by lowering reasoning effort for routine steps and reserving higher effort for planning or difficult decisions.
Read how AI Agents in Action, Second Edition explains reasoning model →
Introduction to Generative AI, Second Edition
A reasoning model is a language model that generates intermediate reasoning steps before producing a final answer, especially for difficult problems.
- How it is used
- Uses intermediate analysis and self-critique before answering constrained questions.
- How deep it goes
- Explains the observable reasoning sequence but stops short of disclosing training specifics.
- Examples
- Illustrates rhyme construction through planning, brainstorming, self-correction, and final poem generation.
- What you can do after
- Evaluate how intermediate reasoning differs from direct answer generation.
Read how Introduction to Generative AI, Second Edition explains reasoning model →
Related concepts
- large language modelkind of · A large language model is a machine-learning model, typically neural, trained on extensive text data to process and generate language.
- language modelkind of · A language model is a computational model that represents patterns in linguistic tokens to estimate sequence probabilities and predict or generate language from context.
- multi-step reasoninguses · Multi-step reasoning derives a conclusion, solution, or decision through multiple connected intermediate inferential or problem-solving steps.
- chain-of-thought promptingcontrasts with · Chain-of-thought prompting is a technique that elicits or guides a model to produce or follow intermediate reasoning steps before a final answer.
These guides are generated from the text of Manning books by Manning's knowledge graph and link to the sections they draw on. The full books are available on liveBook and at manning.com.