Think about it- as I said below, usually it's about things like s vs ss, r vs rr, l vs ll etc

But, even with that interpetation, I don't think it really explains the errors. Like just now I asked how many i's were in "disabilities": https://i.imgur.com/TZFByen.png - it gives the wrong answer and there's no double-i in disabilities to be causing the ambiguity. The follow-on reveals that it generally struggles working with individual letters.

Or, taking a word that is ambigious in this way and adding another of the letter to it to show that it really is just undercounting: https://i.imgur.com/6PaetPK.png

You can also be clear about what you mean and still get the same error: https://i.imgur.com/S0q6vG7.png

Right.

I agree it fails to actually count the letters, but I still think those two interpretations of the question are valid, so making sure the LLM addresses the intended one should be important.

I'm also not sure about the disabilities example: This type of question would be very uncommon I think, not least because there aren't really words with double ii's, and people dont usually ask trivially about the number of characters in a word; rather it's usually about the spelling of a particular syllable (and among those cases, usually 1 vs 2).

Your second example is more convincing; however again we must ask: is it understanding the question right? Because you didnt reset the context and so perhaps it based its last answer on its second last answer in a way that could be valid (inside your last screenshot).

Idk if that makes sense the way I'm trying to explain what I mean.

If I'm understanding, your theory is:

* When asked for r's in strawberry it outputs 2 because it's interpreting your question as whether the second r-sound has one or two r's

* It also miscounts 2 r's in strawberry when you're clear you don't mean that, or use strawberry as part of another phrase, due to how tokens work

But then that makes the part about interpreting the question as "r's in the second r-sound" superfluous (https://i.imgur.com/eYKYRfN.png), if it counts 2 r's in strawberry anyway. There's no need for it to also be misinterpreting what you meant.

Right, but again this would be valid according to the r-sound-interpretation.

To be precise, we actually have 3 possible interpretations of the question:

1- how many r's are spelled in strawberry in total (this is what everyone assumes is the intended question)

2- how many r-sounds are in strawberry

3- is it 1 or 2 r's in strawberry

My theory was just that we don't know which of the three the LLM thinks was intended, given that its answer is valid for 2/3 of them.

But I must admit that it seems hard to find examples where the LLM clearly assumes option 2 or 3.

So maybe the token-based explanation suffices after all.