How are thinking efforts implemented?

Join the discussion

• Hacker News
• they use multitoken prediction behind the scenes, that might interact with the CoT in a strange way. maybe for different thinking modes they have different MTP models? if so thats interesting
  by sometimelurker
• The number of tokens you predict at time (multi or not) has nothing to do with whether the model wants to emit any, some or a lot of reasoning tokens in reasoning tag -- similar to how branch prediction will not really change the for loop iteration count.
  by pyentropy
• At a guess. May be associated with token length context window. Down selecting is consistent with warning message, forcing cutoff to context window. The technical term cache being a synonym. Increasing the headroom for more "thinking" should allow the implementation to access more resources without warning about the cache breaking.
  by __patchbit__
• each model provider does it differently

  from my test with deepseek, when you turn down the thinking effort, the <thinking> block is missing, but the response text includes something looks like the "thinking process", but of course it's less that what were in the thinking block.

  by aynite
• LLMs work by generating the most likely continuation to a prompt. But they can also generate multiple likely continuations. This create multiple branches which in turn can generate even more branches. The LLM can then evaluate the branches, prune the unpromising ones, and merge the best ones. More branches means more tokens, means more effort.
  by bjourne
• this has nothing to do with the thinking effort however
  by simianwords
• It's usually implemented as a keyword the model is trained to react to. In API-only models like Claude you don't have the access to it, as it's inserted into a masked area. The reason the cache might break is that it's inserted at the start, and if you change it the rest will be invalidated.
  by orbital-decay
• I thought it's a loop really... I may be wrong. It's going through a process of thinking, evaluating, outputting, grading, and if grade was low, then loop from thinking again.
  by inthepond
• I suspect it's not possible (as an end user) to get a thinking trace from one of the models. But what happens with "thinking" is that the model has a conversation with itself in an attempt to home in on a better answer to the original prompt.

  The "amount of thinking" is how long this internal conversation is allowed to progress. The longer it goes on the more it costs. It's all part of the token budget but, because this internal dialogue is hidden, it's not obvious to the end user.

  by masfuerte
• > I suspect it's not possible (as an end user) to get a thinking trace from one of the models. But what happens with "thinking" is that the model has a conversation with itself in an attempt to home in on a better answer to the original prompt.

  The model that summarizes what is inside the CoT/|thinking| tags is just an LLM, and it's just as jailbreakable/susceptible to prompt injection as any other LLM: https://x.com/lefthanddraft/status/1991076879877460322 (for those without X; that's Wyatt Walls demonstrating both getting the gemini summarizer to print the raw CoT, as well as just do random calculations, dump its system prompt, etc.)

  by Chu4eeno
• Different models do slight variants.

  Usually it’s done in post training to enforce behavior based on prompt. Ie. System prompt with thinking:max or low or wtv.

  Enforcement then goes via constrained decoding, checking for think token start and end with max lengths, or other variations

  by aabdi
• For Deepseek V4, The main effect of raising the reasoning effort is to add a little section[1] to the end of the system prompt that says "BTW make sure to think really hard! :)"

  If Anthropic's models work the same way, then changing reasoning effort would break the cache because the API has to modify the system prompt given at the very start of the context and rerun the whole thing through the inference server.

  This kind of limitation is one reason Opus 4.8's mid-conversation system messages[2] are actually a pretty big deal (if they actually work).

  [1] https://huggingface.co/deepseek-ai/DeepSeek-V4-Flash/blob/ma...

  [2] https://platform.claude.com/docs/en/build-with-claude/mid-co...

  by Centigonal
• lmao i used to do that manually is that means i raising the effort?
  by firemelt
• I don't get it!? Why not just add the "BTW make sure to think really hard!" at the end in the new message?

  Is it harder to post-train in such a way?

  by simianwords
• > This kind of limitation is one reason Opus 4.8's mid-conversation system messages[2] are actually a pretty big deal (if they actually work).

  Didn't they start injecting system messages telling Claude to calm his tits in overly long and emotional (iirc it triggered on some keywords) chat contexts last year?

  by Chu4eeno
• Take a look at the harmony repo which specifies the internal OpenAI format - the effort level is specified in the context after the <|start|> tag - https://github.com/openai/harmony

  Note that inference libs also have parsers that put hard limits on reasoning tokens with separate counters (similar to how you can put a limit on token generation per completion versus waiting for an <eos>). For that, take a look at vllm reasoning docs.

  by pyentropy
• Examples with inference of different reasoning effort levels is in the OpenAI docs as well - https://developers.openai.com/cookbook/articles/openai-harmo...

  https://docs.vllm.ai/en/latest/features/reasoning_outputs/#a...

  https://developers.openai.com/api/docs/guides/reasoning

  by pyentropy
• Don't work at a lab but I think they might be warping the probability distribution in the decoding step, at least to generate RL examples for training and maybe in production too.

  There aren't other comments discussing this possibility at the moment, but you don't have to take the token predicted as most likely (greedy decoding). Most decoding strategies do something else which is where settings like temperature come in. So if you want the model to "think harder" you can track whether the current tokens are thinking or answer - in OpenAI's system that's called a channel - and then if you're in a thinking block you might get a model output whose top three predictions are:

      60% <|channel=answer|>
      10% Wait, 
      5%  . The
      [...]
  

  Greedy decoding would stop thinking at this point and start answering, but you want the model to keep thinking so you skip that token and select the next most likely which is "Wait, ". The reasoning levels can map to the probability of skipping the channel change tokens.
  by mike_hearn
• I also thought of this as a general idea: intelligence at the sampling level. Broadly you have different tiers of intelligence

  1. take highest probability

  2. based on some light weight code that tracks some state - like number of tokens or some sampling distribution

  3. higher level is using a smaller llm to decide which token to sample (just a thought)

  by simianwords