![\"\"](\"https:\/\/www.newsbeep.com\/us\/wp-content\/uploads\/2025\/07\/Depositphotos_12761044_S-300x200.jpg\"\/)
<\/p>\n <\/p>\n
It turns out that giving AI models time to think before returning answers provides better quality answers. Lilian Weng explains why.<\/p>\n
Imagine if your AI model could pause to think before it speaks. In a recent article by Lilian Weng<\/a>, \u201cWhy We Think,\u201d she explores how giving language models the ability to \u201cthink\u201d at test time (through strategies such as test-time compute and chain-of-thought (CoT) prompting) can lead to significantly better performance. Instead of rushing to produce answers in a single pass, these approaches allow models to spend time reasoning, reflecting, and revising. Inspired by Daniel Kahneman\u2019s dual-process theory of cognition, this shift parallels how humans alternate between fast, instinctive responses and slower, more deliberate thought.<\/p>\n Allowing models to use more computation during inference turns thinking itself into a resource. With CoT prompting, models generate intermediate steps before producing an answer\u2014mirroring how humans solve complex problems. Techniques like beam search, best-of sampling, and process reward models help guide this reasoning process by ranking and selecting the most promising paths.<\/p>\n What\u2019s surprising is that smaller models using test-time strategies can rival much larger ones relying solely on direct decoding. Methods like sequential revision (where a model revises its own response) and reinforcement learning on checkable tasks (like math or code) enable this progress. By optimizing for correctness, these models learn not just to answer but to reason their way to better answers.<\/p>\n