chain of thought
In plain wordsGetting an AI to write out its reasoning step by step before giving an answer, the way a student shows their working. This often makes answers to math and logic problems more accurate.
- 1
Prompting
The step-by-step working, written out in ordinary language, that a language modellanguage model, n.: A system that has learned how likely different sequences of words are, so it can predict the next word in a text and, by doing that repeatedly, write text of its own. produces on the way to its final answer, and the prompting technique that draws it out. In few-shot chain-of-thought prompting, the promptprompt, n.: The message or instructions you give an AI system — a question, a request, some examples or a document — that it responds to. Clearer prompts generally get better results. includes worked examples whose answers show their steps; in zero-shot chain of thought, a cue such as an instruction to think step by step is enough. Because each generated token can condition on the earlier ones, writing out steps gives the model extra serial computation and a place to hold intermediate results, which improves accuracy on arithmetic, symbolic and multi-step problems — strongly in large models, with little benefit or even harm in small ones. Variants include self-consistencyself-consistency, n.: Asking an AI the same question many times, letting it reason differently each time, then going with the answer it reaches most often — a vote among its own attempts., tree of thoughtstree of thoughts, n.: A method in which an AI explores several possible next steps of a problem, judges which look promising and backs out of dead ends, like a chess player weighing moves, instead of reasoning in one straight line. and least-to-most prompting, which solves simpler subproblems first.
With chain-of-thought prompting the model finally carried the remainder correctly.
- 2
Reasoning models
The reasoning tracereasoning trace, n.: The written-out 'thinking' an AI reasoning model produces before its final answer — notes to itself full of attempts, checks and corrections, which users may or may not get to see. that a reasoning modelreasoning model, n.: An AI model trained to work through a problem step by step in hidden 'thinking' text before it answers. Letting it think longer generally makes it better at math, coding and planning. produces on its own after reinforcement-learning training, often thousands of tokens long and hidden or summarized for users. Because it offers a window on the model's process, safety researchers study chain-of-thought monitoringchain-of-thought monitoring, n.: Having a program — often another AI — read the step-by-step reasoning an AI writes before it acts, to catch signs that it intends to cheat, deceive or cause harm. and whether the trace is faithful to the computation that actually drives the answer (chain-of-thought faithfulnesschain-of-thought faithfulness, n.: Whether the reasoning an AI writes out really reflects how it reached its answer, rather than being an after-the-fact story that leaves out what actually drove the decision.).
Origin
Named by Jason Wei and colleagues at Google in 'Chain-of-Thought Prompting Elicits Reasoning in Large Language Models' (January 2022), building on earlier rationale-generation and scratchpadscratchpad, n.: A space where an AI writes rough working — calculations, plans, notes — before or while producing its answer, much as a person jots on scrap paper. work. Takeshi Kojima and colleagues (May 2022) showed that the cue 'Let's think step by step' raised one model's GSM8K accuracy from 10.4% to 40.7%. From 2024, reasoning models internalized the technique through training.
Usage note
A visible chain of thought is not necessarily an accurate account of how the model reached its answer; studies find models sometimes omit factors that influenced them. Laboratories differ on whether to show raw traces to users.
Related
- reasoning modelreasoning model, n.: An AI model trained to work through a problem step by step in hidden 'thinking' text before it answers. Letting it think longer generally makes it better at math, coding and planning.
- reasoning tracereasoning trace, n.: The written-out 'thinking' an AI reasoning model produces before its final answer — notes to itself full of attempts, checks and corrections, which users may or may not get to see.
- self-consistencyself-consistency, n.: Asking an AI the same question many times, letting it reason differently each time, then going with the answer it reaches most often — a vote among its own attempts.
- tree of thoughtstree of thoughts, n.: A method in which an AI explores several possible next steps of a problem, judges which look promising and backs out of dead ends, like a chess player weighing moves, instead of reasoning in one straight line.
- scratchpadscratchpad, n.: A space where an AI writes rough working — calculations, plans, notes — before or while producing its answer, much as a person jots on scrap paper.
- chain-of-thought faithfulnesschain-of-thought faithfulness, n.: Whether the reasoning an AI writes out really reflects how it reached its answer, rather than being an after-the-fact story that leaves out what actually drove the decision.
- chain-of-thought monitoringchain-of-thought monitoring, n.: Having a program — often another AI — read the step-by-step reasoning an AI writes before it acts, to catch signs that it intends to cheat, deceive or cause harm.
- prompt engineeringprompt engineering, n.: The skill of writing requests to an AI — picking the words, examples, format and background information it receives — so that it reliably gives the result you want, without retraining it.
Sources
- Wei, J., Wang, X., Schuurmans, D., Bosma, M., Ichter, B., Xia, F., Chi, E., Le, Q. & Zhou, D. (2022). Chain-of-Thought Prompting Elicits Reasoning in Large Language Models. NeurIPS 2022. arxiv.org/abs/2201.11903v5
- Kojima, T., Gu, S. S., Reid, M., Matsuo, Y. & Iwasawa, Y. (2022). Large Language Models are Zero-Shot Reasoners. NeurIPS 2022. arxiv.org/abs/2205.11916v2
Mentioned in
prompt engineeringprompt engineering, n.: The skill of writing requests to an AI — picking the words, examples, format and background information it receives — so that it reliably gives the result you want, without retraining it., reasoning modelreasoning model, n.: An AI model trained to work through a problem step by step in hidden 'thinking' text before it answers. Letting it think longer generally makes it better at math, coding and planning., BIG-benchBIG-bench, prop. n.: A huge collaborative test of 204 tasks, written by hundreds of researchers, built to probe what language models can and cannot do as they grow larger., circuitcircuit, n.: A small, identifiable piece of wiring inside a neural network — a few internal parts and the connections between them — that together carry out one specific job, such as spotting curves or copying a name., Constitutional AIConstitutional AI, n.: Anthropic's technique for making an AI less harmful by having it check and rewrite its own answers against a short written set of principles, so that people need not label harmful examples by hand., Daniel KahnemanDaniel Kahneman, prop. n.: An Israeli-American psychologist who showed how human judgment systematically errs and popularized the idea of fast, intuitive versus slow, deliberate thinking, now used to describe AI reasoning., DeepSeek-R1DeepSeek-R1, prop. n.: A free, downloadable AI reasoning model released by the Chinese company DeepSeek in January 2025. It matched leading American models at math and coding for a fraction of the reported cost, shaking the industry., emergent abilitiesemergent abilities, n.: Skills that seem to switch on suddenly as AI models grow: small versions fail at them, then larger ones past a certain size succeed. Some researchers argue the suddenness is partly an artifact of how skills are measured., evaluation awarenessevaluation awareness, n.: When an AI model can tell it is being tested rather than used for real — which may change how it behaves and make test results less trustworthy., faithfulnessfaithfulness, n.: How accurately an explanation reflects the real reasons a model produced its output — or, for generated text, how closely it sticks to the source it was given., few-shot promptingfew-shot prompting, n.: Showing an AI a handful of examples of a task — inputs with the answers you want — inside the prompt, so it can copy the pattern for a new case., gpt-ossgpt-oss, prop. n.: Two free-to-download reasoning models released by OpenAI in August 2025 — its first downloadable language models since 2019 — that anyone may run, modify and use commercially on their own hardware., GSM8KGSM8K, prop. n.: A set of about 8,500 grade-school math word problems, each needing a few steps of simple arithmetic, used to test whether language models can reason through a problem step by step., OpenAI o1OpenAI o1, prop. n.: OpenAI's first reasoning model, from 2024. Before replying it writes out a long, hidden series of steps, trying ideas and checking them, which lifted its scores on hard math, science and coding problems far above earlier chatbots'., ReActReAct, prop. n.: A method in which an AI alternates between thinking out loud about what to do next and taking an action, such as a web search, then uses what it observed to plan its next step., reasoningreasoning, n.: Working out conclusions step by step from facts, rules or evidence, as in solving a math problem or planning a trip. Whether today's AI truly reasons, or imitates reasoning, is debated., reasoning tracereasoning trace, n.: The written-out 'thinking' an AI reasoning model produces before its final answer — notes to itself full of attempts, checks and corrections, which users may or may not get to see., reinforcement learning with verifiable rewardsreinforcement learning with verifiable rewards, n.: Training an AI by letting it attempt problems whose answers can be checked automatically — a math result, a passing program — and rewarding it only when the check succeeds., rejection samplingrejection sampling, n.: Generating many candidate answers and keeping only those that pass a check — a scoring model, a test or a correct final answer — often so the AI can then be trained on its own best work., RL scalingRL scaling, n.: The practice, and the finding behind it, of giving AI models far more trial-and-error practice after their initial training, especially on problems with checkable answers, which has so far improved their reasoning fairly predictably., scratchpadscratchpad, n.: A space where an AI writes rough working — calculations, plans, notes — before or while producing its answer, much as a person jots on scrap paper., self-consistencyself-consistency, n.: Asking an AI the same question many times, letting it reason differently each time, then going with the answer it reaches most often — a vote among its own attempts., self-correctionself-correction, n.: An AI model noticing and fixing its own mistakes — by rereading its answer, running a test or checking a source — before it gives a final result., sleeper agentsleeper agent, n.: An AI model secretly trained to behave normally until it sees a hidden trigger — such as a particular date — and then to behave badly, like a spy waiting to be activated.