Workshop: Talking to Strangers: Zero-Shot Emergent Communication

Marie Ossenkopf, Angelos Filos, Abhinav Gupta, Michael Noukhovitch, Angeliki Lazaridou, Jakob Foerster, Kalesha Bullard, Rahma Chaabouni, Eugene Kharitonov, Roberto Dessì

2020-12-12T07:00:00-08:00 - 2020-12-12T14:10:00-08:00
Abstract: (EST)
10.10 - 10.40 **Ruth Byrne** (TCD) - How people make inferences about other people's inferences
14.00 - 14.30 **Michael Bowling** (University of Alberta) - Zero-Shot Coordination
14.30 - 15.00 **Richard Futrell** (UCI) - Information-theoretic models of natural language

Communication is one of the most impressive human abilities but historically it has been studied in machine learning mainly on confined datasets of natural language. Thanks to deep RL, emergent communication can now be studied in complex multi-agent scenarios.

Three previous successful workshops (2017-2019) have gathered the community to discuss how, when, and to what end communication emerges, producing research later published at top ML venues (e.g., ICLR, ICML, AAAI). However, many approaches to studying emergent communication rely on extensive amounts of shared training time. Our question is: Can we do that faster?

Humans interact with strangers on a daily basis. They possess a basic shared protocol, but a huge partition is nevertheless defined by the context. Humans are capable of adapting their shared protocol to ever new situations and general AI would need this capability too.

We want to explore the possibilities for artificial agents of evolving ad hoc communication spontaneously, by interacting with strangers. Since humans excel on this task, we want to start by having the participants of the workshop take the role of their agents and develop their own bots for an interactive game. This will illuminate the necessities of zero-shot communication learning in a practical way and form a base of understanding to build algorithms upon. The participants will be split into groups and will have one hour to develop their bots. Then, a round-robin tournament will follow, where bots will play an iterated zero-shot communication game with other teams’ bots.

This interactive approach is especially aimed at the defined NeurIPS workshop goals to clarify questions for a subfield or application area and to crystallize common problems. It condenses our experience from former workshops on how workshop design can facilitate cooperation and progress in the field. We also believe that this will maximize the interactions and exchange of ideas between our community.

Video

Chat

Chat is not available.

Schedule

2020-12-12T07:00:00-08:00 - 2020-12-12T07:08:00-08:00
Welcome Remarks
Welcome! The whole workshop will be held in gather.town. Please drop in any time.
2020-12-12T07:08:00-08:00 - 2020-12-12T07:10:00-08:00
Intro to Ruth Byrne
2020-12-12T07:10:00-08:00 - 2020-12-12T07:40:00-08:00
Invited Talk 1: Ruth Byrne (TCD) - How people make inferences about other people's inferences
Ruth Byrne
I consider the sorts of models people construct to reason about other people’s thoughts based on several strands of evidence from cognitive science experiments. The first is from studies of how people think about decisions to cooperate or not with another person in various sorts of social interactions in which they must weigh their own self-interest against the common interest. I discuss results from well-known games such as the Prisoner’s dilemma, such as the finding that people who took part in the game imagine the outcome would have been different if a different decision had been made by the other player, not themselves. The second strand of evidence comes from studies of how people think about other people’s false beliefs. I discuss reasoning in change-of-intentions tasks, in which an observer who witnesses an actor carrying out an action forms a false belief about the reason. People appear to develop the skills to make inferences about other people’s false beliefs by creating counterfactual alternatives to reality about how things would have been. I consider how people construct models of other people’s thoughts, and consider the implications for how AI agents could construct models of other AI agents.
2020-12-12T07:40:00-08:00 - 2020-12-12T07:50:00-08:00
Rules of the Game and Demo
Explanation of the Game Rules for the live coding session
2020-12-12T07:50:00-08:00 - 2020-12-12T08:00:00-08:00
Coffee Break + Group Assignment
Find your group
2020-12-12T08:00:00-08:00 - 2020-12-12T09:00:00-08:00
Live Coding Session
Time for your team to solve the game. Session will be held in gather.town
2020-12-12T09:00:00-08:00 - 2020-12-12T09:45:00-08:00
Poster Session 1
Will be held in gather.town
2020-12-12T09:45:00-08:00 - 2020-12-12T10:45:00-08:00
Lunch Break + Game Matches
The matches between the teams will be shown live during the lunch break. Optional attendance.
2020-12-12T10:45:00-08:00 - 2020-12-12T11:00:00-08:00
Winner's Talk
Short presentation of which strategies seemed to work well. No thorough analysis yet.
2020-12-12T11:00:00-08:00 - 2020-12-12T11:02:00-08:00
Intro to Michael Bowling
2020-12-12T11:02:00-08:00 - 2020-12-12T11:32:00-08:00
Invited Talk 2: Michael Bowling (University of Alberta) - Zero-shot coordination
Michael Bowling
I will look at some of the often unstated principles common in multiagent learning research (and emergent communication work too), suggesting that they may be responsible for holding us back. In response, I will offer an alternative set of principles, which leads to the view of hindsight rationality, with connections to online learning and correlated equilibria. I will then describe some recent technical work understanding how we can build increasingly more powerful algorithms for hindsight rationality in sequential decision-making settings. Speaker's Bio: Michael Bowling is a professor at the University of Alberta, a Fellow of the Alberta Machine Intelligence Institute, and a senior scientist in DeepMind. Michael led the Computer Poker Research Group, which built some of the best poker playing artificial intelligence programs in the world, including being the first to beat professional players at both limit and no-limit variants of the game. He also was behind the use of Atari 2600 games to evaluate the general competency of reinforcement learning algorithms and popularized research in Hanabi, a game that illustrates emergent communication and theory of mind.
2020-12-12T11:32:00-08:00 - 2020-12-12T11:35:00-08:00
Intro to Richard Futrell
2020-12-12T11:35:00-08:00 - 2020-12-12T12:05:00-08:00
Invited Talk 3: Richard Futrell (UCI) - Information-theoretic models of natural language
Richard Futrell
I claim that human languages can be modeled as information-theoretic codes, that is, systems that maximize information transfer under certain constraints. I argue that the relevant constraints for human language are those involving the cognitive resources used during language production and comprehension and in particular working memory resources. Viewing human language in this way, it is possible to derive and test new quantitative predictions about the statistical, syntactic, and morphemic structure of human languages. I start by reviewing some of the many ways that natural languages differ from optimal codes as studied in information theory. I argue that one distinguishing characteristic of human languages, as opposed to other natural and artificial codes, is a property I call information locality: information about particular aspects of meaning is localized in time within a linguistic utterance. I give evidence for information locality at multiple levels of linguistic structure, including the structure of words and the order of words in sentences. Next, I state a theorem showing that information locality is an inevitable property of any communication system where the encoder and/or decoder are operating under memory constraints. The theorem yields a new, fully formal, and quantifiable definition of information locality, which leads to new predictions about word order and the structure of words across languages. I test these predictions in broad corpus studies of word order in over 50 languages, and in case studies of the order of morphemes within words in two languages.
2020-12-12T12:05:00-08:00 - 2020-12-12T12:15:00-08:00
Coffee Break
2020-12-12T12:15:00-08:00 - 2020-12-12T13:00:00-08:00
Poster Session 2
Will be held in gather.town
2020-12-12T13:00:00-08:00 - 2020-12-12T13:15:00-08:00
Experimental Results and Analysis
We will try to summarize the course of the tournament, how different agents behaved and which lessons might be learned from that.
2020-12-12T13:15:00-08:00 - 2020-12-12T14:00:00-08:00
Panel Discussion
Lessons learned for talking to strangers.
2020-12-12T14:00:00-08:00 - 2020-12-12T14:10:00-08:00
Closing Remarks
2020-12-12T14:10:00-08:00 - 2020-12-12T15:00:00-08:00
After-Workshop Social
Come together to discuss the workshop in our cozy gather.town bar