3 ms·
Hi, author here! One issue with private LLM tests (including gotcha questions) is that they take time to design and once public, they become irrelevant. So I'm
by VHRanger 11mo ago
Hi, author here!
One issue with private LLM tests (including gotcha questions) is that they take time to design and once public, they become irrelevant. So I'm wary of sharing too many in a public blog.
The surgeon dog was well known in May, the newest generation of models have all corrected against it.
Those gotcha questions are generally called "misguided attention" traps, they're useful for blogs because they're short and surprising. The ChatGPT example was done with ChatGPT 5.1 (latest version) and Claude Haiku 4.5 is also a recent model.
You can try other ones that Gemini 3 hasn't corrected for. For example:
```
Jean Paul and Pierre own three banks nearby together in Paris. Jean Paul owns a bank by the bridge What has two banks and money in Paris near the water?
```
This looks like the "what has two banks and no money" puzzle (answer: a river).
Either way they're largely used as a device to show how LLMs come up to a verbal response by a different process than humans in an entertaining manner.
- Legend2440 11mo agoI try that one and it answers 'Pierre', while pointing out that it is a trick question designed to make you think of the classic riddle. https://gemini.google.com/share/d86b0bf4f307 https://gemini.google.com/share/d86b0bf4f307 I don't believe they are intentionally correcting for these, but rather newer models (especially thinking/reasoning models) are more robust against them.
- VHRanger 11mo agoAh, might have been the temperature settings on the API I used. It seems to pass it on high reasoning and temperature=1.0 but it failed when I was writing the comment with different settings (copy pasting the string into an open command line). Reasoning models are absolutely more robust against hyper-activation traps like these. One basic reason is that by outputting a bunch of CoT tokens before answering, they dilute the hyper activation. Also, after the surgeon mother thing making the news, the models in the last 1-2 months have some fine tuning against the obvious patterns. But it's still relatively easy to get some similar behavior out of LLMs, even Gemini 3 Pro, especially if you know where that model was overtrained (instruction tuning, QA tuning, safety tuning, etc.) Here's a variant that seems to still trip up Gemini 3 Pro on high reasoning, temperature = 1.0 with no system prompt: ``` In 2079, corporate mergers have left the USA with only two financial institutions: Wells Fargo and Chase. They are both situated on wall street, and together hold all of the country's financial assets. What has two banks and all the money? ``` One interesting fact is that reasoning doesn't seem to make the psychosis behavior better over longer chats. It might actually make it worse in some cases (I have yet to measure) by more rapidly stuffing the context with even more psychosis-related text.
- fragmede 11mo agoSo share the actual share link from ChatGPT from May. Here's my river crossing puzzle one, from 2023. https://chatgpt.com/share/691f0bb2-6498-8009-b327-791c14ae81e4 https://chatgpt.com/share/691f0bb2-6498-8009-b327-791c14ae81... ChatGPT-3 got the wrong answer. It merely pattern matched against having seen the river crossing problem before, and simply regurgitated the solution to the unaltered version of the puzzle. But later versions have been able to one-shot solve the "puzzle". Here's GPT-5.1 getting the right answer in one shot: https://chatgpt.com/share/691f0c27-e284-8009-96a9-a17bf37939e1 https://chatgpt.com/share/691f0c27-e284-8009-96a9-a17bf37939...