Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Poker Tournament for LLMs

pokerbattle.ai

Post reply on HN
Print view View original

191–200 of 212 posts

Re: Poker Tournament for LLMs

#191

Post by mh- »

aelaguiz wrote: post #125

This is my area of expertise. I love the experiment. In general games of imperfect information such as Poker, Diplomacy, etc are much much harder than perfect information games such as Chess. Multiplayer (3+) poker in particular is interesting because you cannot achieve a nash equilibrium (e.g. it is not zero sum). That is part of the reason they are a fantastic venue for exploration of the capabilities of LLMs. They…

> pokerskill.com

Cool app, love the concept! Played poker a lot 20 years ago and very little since.

Ran into some minor UX snags (iPhone) - feel free to hit me up if you're looking for feedback.

Re: Poker Tournament for LLMs

#193

Post by dacox »

aelaguiz wrote: post #125

This is my area of expertise. I love the experiment. In general games of imperfect information such as Poker, Diplomacy, etc are much much harder than perfect information games such as Chess. Multiplayer (3+) poker in particular is interesting because you cannot achieve a nash equilibrium (e.g. it is not zero sum). That is part of the reason they are a fantastic venue for exploration of the capabilities of LLMs. They…

Cool app!

I think I found a broken problem (or it’s worded strangely?) and I’m unable to progress beyond it

“Pick 2 that beat villain on board”

“QQ552”

In submitting Queen and 2 to make a full house but it just says

“Incorrect. Two pair on board. Win with a full house or ace kicker”

Re: Poker Tournament for LLMs

#194

Post by solotronics »

eclark wrote: post #135
brrrrrm wrote: post #127

Earlier quoted context omitted.

> None of which are they currently capable what makes you say this? modern LLMs (the top players in this leaderboard) are typically equipped with the ability to execute arbitrary Python and regularly do math + random generations. I agree it's not an efficient mechanism by any means, but I think a fine-tuned LLM could play near GTO for almost all hands in a small ring setting

To play GTO currently you need to play hand ranges. (For example when looking at a hand I would think: I could have AKs-ATs, QQ-99, and she/he could have JT-98s, 99-44, so my next move will act like I have strength and they don't because the board doesn't contain any low cards). We have do this since you can't always bet 4x pot when you have aces, the opponents will always know your hand strength directly. LLM's aren…

If you could, theoretically, make a LLM that could actually excel at poker would that mean that it is good at lying to people?

Re: Poker Tournament for LLMs

#195

Post by sammy2255 »

mpavlov wrote: post #121
sammy2255 wrote: post #94

Whis was built on Vercel and its shitting the bed right now

(author of PokerBattle is here) Well, you're not wrong :) Vercel is not the one to blame here, it's my skill issue. Entire thing was vibecoded by me — product manager with no production dev experience. Not to promote vibecoding, but I couldn't do it myself the other way.

sorry i was mean

Re: Poker Tournament for LLMs

#196

Post by eru »

CamperBob2 wrote: post #185
mrob wrote: post #118

Earlier quoted context omitted.

>No implementations of models you’re talking to today are just raw autorrgressive predictors, taking the most likely next token. Set the temperature to zero and that's exactly what you get. The point is the randomness is something applied externally, not a "core concept" for the LLM.

Set the temperature to zero and that's exactly what you get. In some NN implementations, randomness is actually pretty important to keep the gradients from getting stuck at local minima/maxima. Is that true for LLMs, or is it not something that applies at all?

Are you talking about training?

Re: Poker Tournament for LLMs

#197

Post by eru »

eclark wrote: post #124
_ink_ wrote: post #53

Earlier quoted context omitted.

> LLMs do not have a mechanism for sampling from given probability distributions. They could have a tool for that, tho.

They would need to lie, which they can't currently do. To play at our current best, our approximation of optimal play involves ranges. Thinking about your hand as being any one of a number of cards. Then imagine that you have combinations of those hands, and decide what you would do. That process of exploration by imagination doesn't work with an eager LLM using huge encoded context.

Why would they need to lie? Where's the lying in Poker?

(Ignore for a moment that LLMs can lie just fine.)

What you are describing is exploring a range of counterfactuals. That's not lying.

Re: Poker Tournament for LLMs

#198

Post by eru »

btilly wrote: post #150
michalsustr wrote: post #31

I have PhD in algorithmic game theory and worked on poker. 1) There are currently no algorithms that can compute deterministic equilibrium strategies [0]. Therefore, mixed (randomized) strategies must be used for professional-level play or stronger. 2) In practice, strong play has been achieved with: i) online search and ii) a mechanism to ensure strategy consistency. Without ii) an adaptive opponent can learn to exp…

What you describe is not a contrast to chess. Current LLMs also do not play chess well. Generally they play at the 1000-1300 ELO level. Playing specific games well requires specialized game-specific skills. A general purpose LLM generally lacks those. Future LLMs may be slightly better. But for the foreseeable future, the real increase of playing strength is having an LLM that knows when to call out to external tools…

Well, what would be interesting, is if the LLM came up with its own specific game tools.

Eg you describe your variant of fantasy chess or funny Poker, and it would cobble together some ad hoc code that would help it play that game.

The code wouldn't need to be great from the get go, since the LLM can react to corner cases and errors.

Re: Poker Tournament for LLMs

#199

Post by zie1ony »

48terry wrote: post #159
zie1ony wrote: post #111

Hi there, I'm also working on LLMs in Texas Hold'em :) First of all, congrats on your work. Picking a form of presenting LLMs, that playes poker is a hard task, and I like your approach in presenting the Action Log. I can share some interesting insights from my experiments: - Findin strategies is more interesting than comparing different models. Strategies can get pretty long and specific. For example, if part of the…

Question: What makes LLMs well-suited for the task of poker compared to other approaches?

They are not, and that's the whole point of doing this research. If we can build good benchmark, models developers would have nice goal.

Re: Poker Tournament for LLMs

#200

Post by ngruhn »

rzk wrote: post #29

See also: https://nof1.ai/ Six LLMs were given $10k each to trade in real markets autonomously using only numerical market data inputs and the same prompt/harness.

So the Chinese ones make profit and the silicon valley LLMs are burning money. Sounds about right.
Post reply on HN