Human-Level Play in Diplomacy Combining Language Models with Strategic Reasoning

Re:Human-Level Play in Diplomacy Combining Language Models w... by ekdinan on Tuesday November 22, 2022 @12:46PM (#33708552)

Indeed, consistency (with the lengthy dialogue histories, but also the game state as well as game plans) was a huge challenge for us. We spent a lot of time working on techniques for detecting and filtering these kinds of low quality messages. You can see the Supplementary Materials in the paper for full details, but TL;DR: we built a suite of classifiers for detecting common mistakes (training classifiers to discriminate between human messages and counterfactuals), and used these classifiers as an ensemble which acted as a filter on top of message generation.

[ Parent | on HN ]

Re:Human-Level Play in Diplomacy Combining Language Models w... by O__________O on Tuesday November 22, 2022 @01:31PM (#33709176)

Congrats on getting research published.

Feel like a hack would have been to try to force dialogue into an extractable form that stored a state model relevant to the game, even additional hacks like asking the opposing player to restate their understanding of prior agreements; disclosure that I have no idea how the game Diplomacy works, so might be irrelevant.

Beyond that, no idea how Facebook manages its AI research, but quick Google confirms my memory that Meta/Facebook has done prior research on enabling AI memory capabilities related to recall, forgetting, etc.; which I mention just in case you were not aware.

[ Parent | on HN ]

Re:Human-Level Play in Diplomacy Combining Language Models w... by eightysixfour on Tuesday November 22, 2022 @03:20PM (#33710897)

Do you think it is an easier challenge than than just “converse with a human” since there is a purpose/game state which is driving the interactions?

[ Parent | on HN ]

Re:Human-Level Play in Diplomacy Combining Language Models w... by sanxiyn on Wednesday November 23, 2022 @03:47AM (#33716679)

It is closed domain, so yes, certainly easier than open domain chatbot.

[ Parent | on HN ]

Re:Human-Level Play in Diplomacy Combining Language Models w... by dqpb on Tuesday November 22, 2022 @06:27PM (#33713014)

Did you capture conversational "transactions" as structured data in the game state, or was the chat history itself the only storage for that aspect of the game?

I would think you could avoid much of this issue by creating a more sophisticated structured game model and use the language model only for converting between structured and unstructured data.

[ Parent | on HN ]

Re:Human-Level Play in Diplomacy Combining Language Models w... by sanxiyn on Wednesday November 23, 2022 @03:51AM (#33716712)

They do have structured game model although it doesn't capture everything in chat. Language model still had lots of problems with consistency even with structured game model input.

[ Parent | on HN ]