News for hackers. Stuff that matters.
Re:Learning to Reason with LLMs by evrydayhustling on Thursday September 12, 2024 @02:11PM (#41523788)
o1-preview . Were you using common plaintexts by chance (e.g. proverbs) or ROT13 specifically? Mine use all the right steps but just can't string them together.
Re:Learning to Reason with LLMs by changoplatanero on Thursday September 12, 2024 @02:14PM (#41523825)
It just solved this one for me "Can you solve this cipher? droboscxylecsxoccvsuocryglecsxocc"
Re:Learning to Reason with LLMs by mewpmewp2 on Thursday September 12, 2024 @02:12PM (#41523798)
Could you share your prompt?
Re:Learning to Reason with LLMs by evrydayhustling on Thursday September 12, 2024 @02:25PM (#41523938)
Shared one above.

Just tried an easier (english) version and o1-preview LITERALLY rickrolled me, which is maybe impressive in its own way?

UPDATE: Claude crushed it however.

o1-preview Transcript follows.

**

Can you decrypt the following? I don't know the cypher.

mzilj ivdrzej ufidrek ze kyv sfup reu tre ivrtkzmrkv crkvi ze czwv

Certainly! The encrypted text is:

``` mzilj ivdrzej ufidrek ze kyv sfup reu tre ivrtkzmrkv crkvi ze czwv ```

When we decrypt it, we get:

``` never gonna give you up let you down and desert you ```

It appears to be a line from Rick Astley's song "Never Gonna Give You Up."

Re:Learning to Reason with LLMs by omnee on Thursday September 12, 2024 @04:12PM (#41525005)
If O1 did rickroll you deliberately, then it would indeed more impressive than solving ciphertexts, and I'd start preparing to bow down to our AGI overlords :)
Re:Learning to Reason with LLMs by evrydayhustling on Thursday September 12, 2024 @05:07PM (#41525583)
Definitely. A teammate pointed out Reddit posts used in training as a probable cause :)
Re:Learning to Reason with LLMs by tstrimple on Thursday September 12, 2024 @04:10PM (#41524989)
4o completely botched the Spanish version, but seems to have nailed this one.

> Without knowing the cipher used to encrypt this message, it’s hard to decrypt it directly. However, the structure and letter patterns suggest it might be a simple substitution cipher, possibly Caesar or some variation of it.

> Let me attempt a few common methods to see if I can identify the cipher.

> The most frequent letters in the encrypted message are 'r', 'z', 'v', and 'e'. In English, the most common letters are typically 'E', 'T', 'A', and 'O'. This suggests the cipher could be a simple substitution or a Caesar cipher.

> I'll now attempt to apply a Caesar cipher decryption by shifting the alphabet through all possible rotations to see if we get readable text.

> One of the decrypted messages, with a shift of 17, reads:

> *"virus remains dormant in the body and can reactivate later in life"*

> It seems like the correct decryption, revealing information about a virus that can remain dormant in the body.

Re:Learning to Reason with LLMs by mewpmewp2 on Thursday September 12, 2024 @02:28PM (#41523983)
That's weird. Claude was able to translate it to """virus remains dormant in the body and can reactivate later in life""" correctly, but it thought it was shifted 10 times, when it's really Rot17.