TE
TechEcho
Home24h TopNewestBestAskShowJobs
GitHubTwitter
Home

TechEcho

A tech news platform built with Next.js, providing global tech news and discussions.

GitHubTwitter

Home

HomeNewestBestAskShowJobs

Resources

HackerNews APIOriginal HackerNewsNext.js

© 2025 TechEcho. All rights reserved.

Reasoning Models Can Be Effective Without Thinking

21 pointsby mfiguiereabout 1 month ago

2 comments

nsoonhuiabout 1 month ago
I&#x27;m not entirely sure how this kind of study jives well with other study, such as &quot;Reasoning models don&#x27;t always say what they think&quot; [0], discussion [1].<p>To quote the article:<p><pre><code> We can’t be certain of either the “legibility” of the Chain-of-Thought (why, after all, should we expect that words in the English language are able to convey every single nuance of why a specific decision was made in a neural network?) or its “faithfulness”—the accuracy of its description. There’s no specific reason why the reported Chain-of-Thought must accurately reflect the true reasoning process; there might even be circumstances where a model actively hides aspects of its thought process from the user. </code></pre> So if we can&#x27;t trust the reasoning, then what&#x27;s the point of checking whether they are &quot;effective&quot; or not?<p>[0]: <a href="https:&#x2F;&#x2F;www.anthropic.com&#x2F;research&#x2F;reasoning-models-dont-say-think" rel="nofollow">https:&#x2F;&#x2F;www.anthropic.com&#x2F;research&#x2F;reasoning-models-dont-say...</a><p>[1]: <a href="https:&#x2F;&#x2F;news.ycombinator.com&#x2F;item?id=43572374">https:&#x2F;&#x2F;news.ycombinator.com&#x2F;item?id=43572374</a>
skeritabout 1 month ago
&gt; When controlling for the number of tokens, NoThinking outperforms Thinking across a diverse set of seven challenging reasoning datasets<p>Interesting. I thought the &quot;thinking&quot; was useful because it pulls in a lot of concepts into the context, but I guess not then?<p>It has also been said before that the text a model outputs during its Thinking step isn&#x27;t actually a view into its inner thoughts. There are times when the model will think X but eventually answer Y.<p>But even so: the models _are_ better, right? So is the Thinking step then mostly useful during training?