Zobrazeno 1 - 10
of 560
pro vyhledávání: '"Daw, Nathaniel D"'
Publikováno v:
Computational Psychiatry, Vol 4, Pp 1-17 (2020)
Anxiety disorders are characterized by a range of aberrations in the processing of and response to threat, but there is little clarity what core pathogenesis might underlie these symptoms. Here we propose that a particular set of unrealistically pess
Externí odkaz:
https://doaj.org/article/4368a93cbb864d22a9a94d7c3eb82051
Typical models of learning assume incremental estimation of continuously-varying decision variables like expected rewards. However, this class of models fails to capture more idiosyncratic, discrete heuristics and strategies that people and animals a
Externí odkaz:
http://arxiv.org/abs/2402.16668
Autor:
Correa, Carlos G., Sanborn, Sophia, Ho, Mark K., Callaway, Frederick, Daw, Nathaniel D., Griffiths, Thomas L.
Human behavior is often assumed to be hierarchically structured, made up of abstract actions that can be decomposed into concrete actions. However, behavior is typically measured as a sequence of actions, which makes it difficult to infer its hierarc
Externí odkaz:
http://arxiv.org/abs/2311.18644
Human behavior emerges from planning over elaborate decompositions of tasks into goals, subgoals, and low-level actions. How are these decompositions created and used? Here, we propose and evaluate a normative framework for task decomposition based o
Externí odkaz:
http://arxiv.org/abs/2211.03890
Autor:
Kumar, Sreejan, Correa, Carlos G., Dasgupta, Ishita, Marjieh, Raja, Hu, Michael Y., Hawkins, Robert D., Daw, Nathaniel D., Cohen, Jonathan D., Narasimhan, Karthik, Griffiths, Thomas L.
Strong inductive biases give humans the ability to quickly learn to perform a variety of tasks. Although meta-learning is a method to endow neural networks with useful inductive biases, agents trained by meta-learning may sometimes acquire very diffe
Externí odkaz:
http://arxiv.org/abs/2205.11558
Autor:
Kumar, Sreejan, Dasgupta, Ishita, Daw, Nathaniel D., Cohen, Jonathan D., Griffiths, Thomas L.
The ability to acquire abstract knowledge is a hallmark of human intelligence and is believed by many to be one of the core differences between humans and neural network models. Agents can be endowed with an inductive bias towards abstraction through
Externí odkaz:
http://arxiv.org/abs/2204.01437
Autor:
Correa, Carlos G., Sanborn, Sophia, Ho, Mark K., Callaway, Frederick, Daw, Nathaniel D., Griffiths, Thomas L.
Publikováno v:
In Cognition February 2025 255
Autor:
Velázquez-Vargas, Carlos A.1 (AUTHOR) cavargas@princeton.edu, Daw, Nathaniel D.1,2 (AUTHOR), Taylor, Jordan A.1,2 (AUTHOR)
Publikováno v:
PLoS Computational Biology. 9/27/2024, Vol. 20 Issue 9, p1-43. 43p.
Autor:
Jahn, Caroline I., Markov, Nikola T., Morea, Britney, Daw, Nathaniel D., Ebitz, R. Becket, Buschman, Timothy J.
Publikováno v:
In Cell 14 March 2024 187(6):1476-1489
Autor:
Kumar, Sreejan, Dasgupta, Ishita, Cohen, Jonathan D., Daw, Nathaniel D., Griffiths, Thomas L.
In recent years, meta-learning, in which a model is trained on a family of tasks (i.e. a task distribution), has emerged as an approach to training neural networks to perform tasks that were previously assumed to require structured representations, m
Externí odkaz:
http://arxiv.org/abs/2010.02317