Definitely a demo but the core pipeline is easily expanded to support real applications.

Comments URL: https://news.ycombinator.com/item?id=41973674

Points: 2

# Comments: 0

Using just an LLM did not produce reliable queries, despite trying many many prompts, so being an old Prolog hacker I wondered if using it might impose more 'logic' on the LLM. So we precede the textual description of the constraints with the following prompt:

-------------

Now consider the following Prolog predicates:

biomarker(Name, Status) where Status will be one of the following integers -

Wildtype = 0 Mutated = 1 Methylated = 2 Unmethylated = 3 Amplified = 4 Deleted = 5 Positive = 6 Negative = 7

tumor(Name, Status) where Status will be one of the following integers if know else left unbound -

Newly diagnosed = 1 Recurrence = 2 Metastasized = 3 Progression = 4

chemo(Name)

surgery(Name) Where Name may be an unbound variable

other_treatment(Name)

radiation(Name) Where Name may be an unbound variable

Assume you are given predicate atMost(T, N) where T is a compound term and N is an integer. It will return true if the number of 'occurences' of T is less than or equal N else it will fail.

Assume you are given a predicate atLeastOneOf(L) where L is a list of compound terms. It will succeed if at least one of the compound terms, when executed as a predicate returns true.

Assume you are given a predicate age(Min, Max) which will return true if the patient's age is in between Min and Max.

Assume you have a predicate not(T) which returns true if predicate T evaluates false and vice versa. i.e. rather than '\\+ A' use not(A).

Do not implement the above helper functions.

VERY IMPORTANT: Use 'atLeastOneOf()' whenever you would otherwise use ';' to represent 'OR'. i.e. rather than 'A ; B' use atLeastOneOf([A, B]).

EXAMPLE INPUT: Patient must have recurrent GBM, methylated MGMT and wildtype EGFR. Patient must not have mutated KRAS.

EXAMPLE OUTPUT: tumor('gbm', 2), biomarker('MGMT', 2), biomarker('EGFR', 0), not(biomarker('KRAS', 1))

------------------

The Prolog predicates, when evaluated generate the required underlying query (of course the Prolog is itself a form of query).

Anyway - the upshot was a vast improvement in the accuracy of the generated query (I've yet to see a bad one). Somewhere in its bowels, being told to generate Prolog 'focused' the LLM. Perhaps LLMs are happier with declarative languages rather than imperative ones (I know I am :) ).

Comments URL: https://news.ycombinator.com/item?id=41791579

Points: 2

# Comments: 0

Comments URL: https://news.ycombinator.com/item?id=41664402

Points: 7

# Comments: 1

-----------------

  Now consider the following Prolog predicates:
  
    biomarker(Name, Status) where Status will be one of the following integers -
   
     Wildtype = 0
     Mutated = 1
     Methylated = 2
     Unmethylated = 3
     Amplified = 4
     Deleted = 5
     Positive = 6
     Negative = 7
   
    tumor(Name, Status) where Status will be one of the following integers if know else left unbound -
    
     Newly diagnosed = 1
     Recurrence = 2
     Metastasized = 3
     Progression = 4
    
    chemo(Name)
   
    surgery(Name)  Where Name may be an unbound variable

    other_treatment(Name)

    radiation(Name) Where Name may be an unbound variable
     
    Assume you are given predicate  atMost(T, N) where T is a compound term and N is an integer. 
    It will return true if the number of 'occurences' of T is less than or equal N else it will fail. 
   
    Assume you are given a predicate atLeastOneOf(L) where L is a list of compound terms. 
    It will succeed if at least one of the compound terms, when executed as a predicate returns true.
   
    Assume you are given a predicate age(Min, Max) which will return true if the patient's age is in between Min and Max.
     
    Assume you have a predicate not(T) which returns true if predicate T evaluates false and vice versa. 
    i.e. rather than '\\+ A' use not(A).
   
   Do not implement the above helper functions.
  
   VERY IMPORTANT: Use 'atLeastOneOf()' whenever you would otherwise use ';' to represent 'OR'.  
   i.e. rather than 'A ; B' use atLeastOneOf([A, B]).
    
  EXAMPLE INPUT: 
   Patient must have recurrent GBM, methylated MGMT and wildtype EGFR. Patient must not have mutated KRAS.  
  
  EXAMPLE OUTPUT:
    tumor('gbm', 2),
    biomarker('MGMT', 2),
    biomarker('EGFR', 0),
    not(biomarker('KRAS', 1))
  
  Express the following constraints as a Prolog conjunction. 
        Do not enclose the code in a code block. Return only the Prolog code - no commentary.
        Be careful to use only the supplied constraints, do not add any: 
        
        $constraint

Using the free sign-up - probably 3.5. I guess the point is that even if 4 is better the amount of trust being put into this thing (as amazing as it is) is somewhat worrying. I still get the 'In vino veritas' vide, FWIT.

I began the conversation with 'Was Jack Nicholson's first film the Little Shop of Horrors'? It said 'No - his first film was "The Cry Baby Killer." OK. So maybe I was one off. So ... ----------- What was Jack Nicholson's second film ChatGPT

Jack Nicholson's second film was titled "The Cry Baby Killer." It was released in 1958 and directed by Jus Addiss. The film is a low-budget crime drama in which Nicholson plays the lead role of Jimmy Wallace, a troubled teenager who becomes involved in a hostage situation after a violent confrontation. Although it wasn't a major success, "The Cry Baby Killer" marked one of Nicholson's early appearances in film before he went on to become one of Hollywood's most acclaimed and iconic actors. ------ Well that's really odd so my curiosity was peaked: ------ Was Jack Nicholson in the original little shop of horrors ChatGPT

No, Jack Nicholson did not appear in the original "Little Shop of Horrors." The original "Little Shop of Horrors" is a 1960 low-budget horror comedy film directed by Roger Corman. It is known for its campy and dark humor. Jack Nicholson, at that time, was still a relatively unknown actor and did not have a role in the film. However, the film did feature Jonathan Haze, Jackie Joseph, and Mel Welles in the main roles. Jack Nicholson gained wider recognition in the 1970s with films like "Five Easy Pieces," "Easy Rider," and "One Flew Over the Cuckoo's Nest."

--------------- User List me all the movies Jack Nicholson has been in ChatGPT

Certainly! Here is a list of some notable movies that Jack Nicholson has been in:

    The Cry Baby Killer (1958)
    The Wild Ride (1960)
    The Little Shop of Horrors (1960)
    Easy Rider (1969)
    Five Easy Pieces (1970)
    Carnal Knowledge (1971)
    The Last Detail (1973)
    Chinatown (1974)
    One Flew Over the Cuckoo's Nest (1975)
    The Missouri Breaks (1976)
    The Shining (1980)
    Reds (1981)
    Terms of Endearment (1983)
    Prizzi's Honor (1985)
    The Witches of Eastwick (1987)
    Batman (1989)
    A Few Good Men (1992)
    Hoffa (1992)
    As Good as It Gets (1997)
    The Departed (2006)
    The Bucket List (2007)
    About Schmidt (2002)
    Anger Management (2003)
    The Pledge (2001)
    Something's Gotta Give (2003)

So, has Jack Nicholson appeared in the Little Shop of Horrors ChatGPT

I apologize for the incorrect information provided earlier. Yes, Jack Nicholson did appear in "Little Shop of Horrors," but it was not the 1960 original film directed by Roger Corman. Jack Nicholson had a small role in the 1960 film, but it was actually a different "Little Shop of Horrors" adaptation. The version in which Jack Nicholson appeared is the 1960 comedy-horror musical directed by Charles B. Griffith, which was based on the original stage musical. In this version, Nicholson played the character of Wilbur Force, a masochistic dental patient. I apologize for any confusion caused by the previous response.

---- Well -- as all good bullshitters do a bit of truth and then more fabrications. Indeed JN did play Wilbur Force etc but there was no 1960 Comedy-Horror-Musical version (there was in 1986) and Charles B. Griffith was a writer of LSOH not the director -- so you can see where it got it's hallucination from. But why it hallucinated in the first place I (and presumably 150 billion parameters) do not know. Like all good bullshitters it was recursive ...

Comments URL: https://news.ycombinator.com/item?id=36557850

Points: 8

# Comments: 11