Public preview
Force Explicit Chain-of-Thought
Add "think step-by-step inside <thinking> tags before answering" — boosts complex-problem accuracy 30-40%.
Goal
Force a language model into "extended thinking" before answering, instead of jumping to output.
Steps
-
1
Pick a question requiring 3+ logical steps (calc, decision, compare).
Expected Outcome
Immediate
Complex-problem accuracy rises 30-40%.
Evidence base
School: Chain-of-Thought Prompting
Founders: Wei et al. · Anthropic (2022)
Wei et al. 2022 "Chain-of-Thought Prompting Elicits Reasoning" proved forcing reasoning-then-answer raises math accuracy from 18% to 57%.
Keywords
Frequently asked questions
How long does "Force Explicit Chain-of-Thought" take?
This exercise takes about 12 minutes, practiced solo in a Creative format.
When will I notice the effect of "Force Explicit Chain-of-Thought"?
Complex-problem accuracy rises 30-40%. In the short term: You learn to diagnose model errors instead of fully accepting or rejecting.
Is "Force Explicit Chain-of-Thought" evidence-based?
Yes — it draws on Chain-of-Thought Prompting (Wei et al., Anthropic (2022)). Wei et al. 2022 "Chain-of-Thought Prompting Elicits Reasoning" proved forcing reasoning-then-answer raises math accuracy from 18% to 57%.
Is "Force Explicit Chain-of-Thought" suitable for beginners?
Its difficulty level is: Intermediate. No external tools required — it can be practiced directly inside the app.
Subscribe for full access
Subscribe to Plus for full access to this practice and the whole catalogue, or try first — free, no sign-in, no commitment.
7 complete practices (one per domain) are free forever, no sign-in.