Skip to main content
While LLMs are generally capable of performing reasoning tasks, they’re probabilistic models that rely on pattern matching from training data. When you’re dealing with complex or edge-case problems, the model might hallucinate or produce logically inconsistent results because it lacks the specific reasoning pathway in its training. However, you can dramatically improve reasoning quality by forcing the model to externalize its thinking process, essentially making it “show its work.” Here are three proven techniques to get LLMs to reason more systematically:
  1. Direct Step-by-Step Instructions: Add “Think step by step” to your prompt. This simple phrase triggers the model to break down complex problems into logical components rather than jumping to conclusions.
  2. Provide a Logical Framework: Instead of hoping the model figures out the right approach, give it the exact framework to follow. This reduces reasoning errors by constraining the solution path. Example: Vague prompt:
    Analyze the impact of climate change on polar bear populations.     Structured prompt:
    Analyze the impact of climate change on polar bear populations using this framework:
    1. Current polar bear population status
    2. Climate change factors affecting Arctic habitat
    3. Direct impacts (habitat loss, hunting changes)
    4. Indirect impacts (food chain disruption)
    5. Future population projections
  3. XML Tags for Process Separation: Use <thinking></thinking> and <answer></answer> tags to separate reasoning from final output. This prevents the model from mixing its working process with the polished result, especially useful for complex multi-step problems. You find our section section about XML tags here.