Chain of thoughtBuilding with AI

Chain-of-thought prompting

4 min readintermediateUpdated 28 Sept 2026
1 · In one line

Chain-of-thought prompting asks for or demonstrates intermediate reasoning text before a final answer, but that text is not a guaranteed view of hidden model internals.

1 · What it is

Chain-of-thought prompting puts intermediate reasoning into generated text. The original pattern showed a model several worked solutions, each with steps before the answer, then asked a new question in the same format. Later work showed that a short step-by-step cue could also elicit intermediate text without examples.

This technique concerns visible tokens: words, equations, or subgoals emitted before a final answer. It should not be confused with neural activations or a provider’s private reasoning trace. Some hosted reasoning models spend tokens on an internal process that the provider does not expose. An API may return a summary of that reasoning, but the summary is still not the raw trace.

There is another limit: a convincing rationale may not faithfully identify what caused the prediction. In one study, researchers nudged models by reordering multiple-choice options so the expected letter was always (A). The nudge pulled the answers, yet the written explanations routinely left it out. Treat reasoning text as a working surface. Check calculations, run code, inspect sources, or use a deterministic verifier when correctness matters.

2 · Why it exists

Many problems break into smaller steps, while a bare answer hides where an error entered.

Problems have partsA multi-step problem can be split into intermediate steps instead of being answered in one jump.
Answers hide errorsA final choice alone does not show which explicit step should be checked.
Explanations can misleadGenerated reasoning can sound plausible while omitting the feature that actually influenced the prediction.
3 · How it works

Separate visible reasoning text from private model computation.

Visible chain-of-thought and private model reasoning are separate A question and worked examples enter a highlighted generation stage that produces visible steps and an answer. A separate private provider lane leads only to an answer or summary. PROMPTED, VISIBLE PATH QUESTION + EXAMPLES3 boxes hold 8 pens each.5 pens are sold.How many are left?show intermediate steps KEY MECHANISMGenerate reasoning textintermediate tokens → final answerthis is produced output VISIBLE OUTPUT3 × 8 = 2424 − 5 = 19Answer: 19 HOSTED REASONING MODEL PROVIDER BOUNDARYprivate reasoning tokens / internal computationnot the same object as the visible explanation above RETURNEDanswer or summarynot a raw private trace Verification target: calculations, sources, code, and outcomes; not confidence in the prose.
Prompted reasoning text is generated output. Hosted reasoning models may keep internal reasoning private and return only an answer or summary.
  1. 1 · showThe prompt can include worked examples whose answers contain intermediate reasoning steps.
  2. 2 · askA new multistep problem follows the same input-and-answer pattern.
  3. 3 · reasonThe model generates intermediate text before committing to the final answer.
  4. 4 · verifyA person or program checks the answer and any externally verifiable steps instead of treating the prose as proof.

Visible chain-of-thought is text the model produced; private reasoning tokens and neural activations are different things.

4 · Where it's used
WhoWhat they askWhat it works with
Math tutor“Where did this solution go wrong?”Generated steps and independently checked calculations
Analyst“Which constraints rule out each option?”A concise justification paired with the final choice
Developer“Why does this test case fail?”An explicit, testable decomposition of the problem
Evaluator“Did the explanation use the evidence it claims?”Counterfactual tests of the generated rationale
5 · What it solves, and what it doesn't
solves
  • Few-shot chain-of-thought demonstrations improved results in the original paper on arithmetic, commonsense, and symbolic reasoning tasks.
  • Intermediate text can make individual calculations or assumptions easier to inspect.
  • A reasoning summary can outline how a response was reached without exposing the raw reasoning tokens.
doesn't solve
  • A fluent explanation is not proof that it faithfully reports the cause of the model's answer.
  • Providers may keep raw reasoning traces hidden and expose only a summary or final answer.
  • Reasoning text does not make a wrong premise, missing fact, or unchecked calculation correct.