TY - JOUR
T1 - Question asking as program generation
AU - Rothe, Anselm
AU - Lake, Brenden M.
AU - Gureckis, Todd M.
N1 - Funding Information:
We thank Chris Barker, Sam Bowman, Noah Goodman, and Doug Markant for feedback and advice. This research was supported by NSF grant BCS-1255538, the John Templeton Foundation Varieties of Understanding project, a John S. McDonnell Foundation Scholar Award to TMG, and the Moore-Sloan Data Science Environment at NYU.
Publisher Copyright:
© 2017 Neural information processing systems foundation. All rights reserved.
PY - 2017
Y1 - 2017
N2 - A hallmark of human intelligence is the ability to ask rich, creative, and revealing questions. Here we introduce a cognitive model capable of constructing humanlike questions. Our approach treats questions as formal programs that, when executed on the state of the world, output an answer. The model specifies a probability distribution over a complex, compositional space of programs, favoring concise programs that help the agent learn in the current context. We evaluate our approach by modeling the types of open-ended questions generated by humans who were attempting to learn about an ambiguous situation in a game. We find that our model predicts what questions people will ask, and can creatively produce novel questions that were not present in the training set. In addition, we compare a number of model variants, finding that both question informativeness and complexity are important for producing human-like questions.
AB - A hallmark of human intelligence is the ability to ask rich, creative, and revealing questions. Here we introduce a cognitive model capable of constructing humanlike questions. Our approach treats questions as formal programs that, when executed on the state of the world, output an answer. The model specifies a probability distribution over a complex, compositional space of programs, favoring concise programs that help the agent learn in the current context. We evaluate our approach by modeling the types of open-ended questions generated by humans who were attempting to learn about an ambiguous situation in a game. We find that our model predicts what questions people will ask, and can creatively produce novel questions that were not present in the training set. In addition, we compare a number of model variants, finding that both question informativeness and complexity are important for producing human-like questions.
UR - http://www.scopus.com/inward/record.url?scp=85046999920&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85046999920&partnerID=8YFLogxK
M3 - Conference article
AN - SCOPUS:85046999920
VL - 2017-December
SP - 1047
EP - 1056
JO - Advances in Neural Information Processing Systems
JF - Advances in Neural Information Processing Systems
SN - 1049-5258
T2 - 31st Annual Conference on Neural Information Processing Systems, NIPS 2017
Y2 - 4 December 2017 through 9 December 2017
ER -