Hacker News Reader

Comment by bandrami | Hacker News Reader

Wait, no, that's the category error I'm talking about. Any answer other than "that was the most likely next token given the context" is untrue. It is not describing what actually happened.

View on HN

I think this statement is on the same level as "a human cannot explain why they gave the answer they gave because they cannot actually introspect the chemical reactions in their brain." That is true, but a human often has an internal train of thought that preceded their ultimate answer, and it is interesting to know what that train of thought was.

In the same way, it is often quite instructive to know what the reasoning trace was that preceded an LLM's answer, without having to worry about what, mechanically, the LLM "understood" about the tokens, if this is even a meaningful question.

But it's not a reasoning trace. Models could produce one if they were designed to (an actual stack of the calls and the states of the tensors with each call, probably with a helpful lookup table for the tokens) but they specifically haven't been made to do that.

When you put an LLM in reasoning mode, it will approximately have a conversation with itself. This mimics an inner monologue.

That conversation is held in text, not in any internal representation. That text is called the reasoning trace. You can then analyse that trace.

Unless things have changed drastically in the last 4 months (the last time I looked at it) those traces are not stored but reconstructed when asked. Which is still the same problem.

For a typical coding agent, there are intermediate tool call outputs and LLM commentary produced while it works on a task and passed to the LLM as context for follow up requests. (Hence the term agent: it is an LLM call in a loop.) You can easily see this with e.g. Claude Code, as it keeps track of how much space is left in the context and requires "context compaction" after the context gradually fills up over the course of a session.

In this regard, the reasoning trace of an agent is trivially accessible to clients, unlike the reasoning trace of an individual LLM API call; it's a higher level of abstraction. Indeed, I implemented an agent just the other day which took advantage of this. The OP that you originally replied to was discussing an agentic coding process, not an individual LLM API call.

Well, right, I see those reasoning stages in reasoning models with Ollama and if you ask it what its reasoning was after the fact what it says is different than what it said at the time.

I can't speak to your specific set up, but it sounds like you're halfway there if you can access the previous traces? All anyone can ask for is "show me the traces that led up to this point"; the "why did you do this" is a notational convenience for querying that data. If your set up isn't summarizing those traces correctly, then that sounds like a specific bug in the context or model quality, but the point is that the traces exist and are queryable in the first place, however you choose to do that.

(I am still primarily talking about agent traces, like the original OP, not internal reasoning blocks for a particular LLM call, though - which may or may not be available in context afterwards.)

In particular, asking "why" isn't a category error here, although there's only a meaningful answer if the model has access to the previous traces in its context, which is sometimes true and sometimes not.

They aren't necessarily "stored" but they are part of the response content. They are referred to as reasoning or thinking blocks. The big 3 model makers all have this in their APIs, typically in an encrypted form.

Reconstruction of reasoning from scratch can happen in some legacy APIs like the OpenAI chat completions API, which doesn't support passing reasoning blocks around. They specifically recommend folks to use their newer esponses API to improve both accuracy and latency (reusing existing reasoning).

There can be higher- and lower-level descriptions of the same phenomenon. when the kettle boils, it’s because the water molecules were heated by the electric element, but it’s also because I wanted a cup of tea.

the llm has no wants

If the reason the LLM retroactively invents for it's previous mistakes is still useful for getting the LLM to not make that kind of mistake again, then the distinction you're driving at doesn't matter.

> Any answer other than "that was the most likely next token given the context" is untrue.

"Because the matrix math resulted in the set of tokens that produced the output". "Because the machine code driving the hosting devices produced the output you saw". "Because the combination of silicon traces and charges on the chips at that exact moment resulted in the output". "Because my neurons fired in a particular order/combination".

I don't see how your statement is any more useful. If an LLM has access to reasoning traces it can realistically waddle down the CoT and figure out where it took a wrong turn.

Just like a human does with memories in context - does't mean that's the full story - your decision making is very subconscious and nonverbal - you might not be aware of it, but any reasoning you give to explain why you did something is bound to be an incomplete story, created by your brain to explain what happened based on what it knows - but there's hidden state it doesn't have access to. And yet we ask that question constantly.

well, do you want something useful or something true?

the word why is used to get something true.

If you want to be pedantic about it you could phrase it as follows.

When the LLM was in reasoning mode, in the reasoning context it often expressed statement X. Given that, and the relevance of statement X to the taken action. It seems likely that the presence of statement X in the context contributed to this action. Besides, the presence of statement X in the reasoning likely means that given the previous context embeddings of X are close to the context.

Hence we think that the action was taken due to statement X.

And that output could have come from an LLM introspecting it's own reasoning.

I don't think that phrasing things so pedanticaly is worth the extra precision though. Especially not for the statement that inspecting the reasoning logs of sn LLM can help give insight on why an LLM acted a certain way.

Reply on news.ycombinator.com