diff --git a/content/publication/cao2023semantics.md b/content/publication/cao2023semantics.md index f982904..99104c1 100644 --- a/content/publication/cao2023semantics.md +++ b/content/publication/cao2023semantics.md @@ -8,10 +8,10 @@ title = "A Semantics for Causing, Enabling, and Preventing Verbs Using Structural Causal Models" date = "2023-05-10" -authors = ["A. Cao","A. Geiger","E. Kreiss","T. Icard","T. Gerstenberg"] +authors = ["A. Cao\\*","A. Geiger\\*","E. Kreiss\\*","T. Icard","T. Gerstenberg"] publication_types = ["3"] publication_short = "_Proceedings of the 45th Annual Conference of the Cognitive Science Society_" -publication = "Cao A., Geiger A., Kreiss E., Icard T., Gerstenberg T. (2023). A Semantics for Causing, Enabling, and Preventing Verbs Using Structural Causal Models. In _Proceedings of the 45th Annual Conference of the Cognitive Science Society_." +publication = "Cao A.\\*, Geiger A.\\*, Kreiss E.\\*, Icard T., Gerstenberg T. (2023). A Semantics for Causing, Enabling, and Preventing Verbs Using Structural Causal Models. In _Proceedings of the 45th Annual Conference of the Cognitive Science Society_." abstract = "When choosing how to describe what happened, we have a number of causal verbs at our disposal. In this paper, we develop a model-theoretic formal semantics for nine causal verbs that span the categories of CAUSE, ENABLE, and PREVENT. We use structural causal models (SCMs) to represent participants' mental construction of a scene when assessing the correctness of causal expressions relative to a presented context. Furthermore, SCMs enable us to model events relating both the physical world as well as agents' mental states. In experimental evaluations, we find that the proposed semantics exhibits a closer alignment with human evaluations in comparison to prior accounts of the verb families." image_preview = "" selected = false diff --git a/content/publication/gershman2016goals.md b/content/publication/gershman2016goals.md index c9a565a..e6b0e20 100644 --- a/content/publication/gershman2016goals.md +++ b/content/publication/gershman2016goals.md @@ -9,10 +9,10 @@ title = "Plans, habits, and theory of mind" date = "2016-01-01" -authors = ["S. J. Gershman ","T. Gerstenberg", "C. L. Baker", "F. Cushman"] +authors = ["S. J. Gershman","T. Gerstenberg", "C. L. Baker", "F. Cushman"] publication_types = ["2"] publication_short = "_PLoS ONE_" -publication = "Gershman*, S. J., Gerstenberg*, T., Baker, C. L., & Cushman, F. (2016). Plans, habits, and theory of mind. _PLoS ONE_, 11(9), e0162246. " +publication = "Gershman, S. J.\\*, Gerstenberg, T.\\*, Baker, C. L., & Cushman, F. (2016). Plans, habits, and theory of mind. _PLoS ONE_, 11(9), e0162246. " image_preview = "" selected = false projects = [] diff --git a/content/publication/mccoy2012probabilistic.md b/content/publication/mccoy2012probabilistic.md index 474a493..05df6e2 100644 --- a/content/publication/mccoy2012probabilistic.md +++ b/content/publication/mccoy2012probabilistic.md @@ -9,10 +9,10 @@ title = "Why blame Bob? Probabilistic generative models, counterfactual reasoning, and blame attribution" date = "2012-01-01" -authors = ["J. McCoy* ","T. D. Ullman* ","A. Stuhlmüller","T. Gerstenberg","J. B. Tenenbaum"] +authors = ["J. McCoy\\*","T. D. Ullman\\*","A. Stuhlmüller","T. Gerstenberg","J. B. Tenenbaum"] publication_types = ["3"] publication_short = "_Proceedings of the 34th Annual Conference of the Cognitive Science Society_" -publication = "McCoy, J.*, Ullman, T.*, Stuhlmüller, A., Gerstenberg, T., & Tenenbaum, J. B. (2012). Why blame Bob? Probabilistic generative models, counterfactual reasoning, and blame attribution. In _Proceedings of the 34th Annual Conference of the Cognitive Science Society_ (pp. 1996-2001). Austin, TX: Cognitive Science Society." +publication = "McCoy, J.\\*, Ullman, T.\\*, Stuhlmüller, A., Gerstenberg, T., & Tenenbaum, J. B. (2012). Why blame Bob? Probabilistic generative models, counterfactual reasoning, and blame attribution. In _Proceedings of the 34th Annual Conference of the Cognitive Science Society_ (pp. 1996-2001). Austin, TX: Cognitive Science Society." image_preview = "" selected = false projects = [] diff --git a/content/publication/outa2022stop.md b/content/publication/outa2022stop.md index 3068ae5..5bc5a21 100644 --- a/content/publication/outa2022stop.md +++ b/content/publication/outa2022stop.md @@ -8,10 +8,10 @@ title = "Stop, children what's that sound? Multi-modal inference through mental simulation" date = "2022-05-12" -authors = ["J. Outa","X. Jia Zhou","H. Gweon","T. Gerstenberg"] +authors = ["J. Outa\\*","X. Jia Zhou\\*","H. Gweon","T. Gerstenberg"] publication_types = ["3"] publication_short = "_Cognitive Science Proceedings_" -publication = "Outa* J., Zhou* X. J., Gweon H., Gerstenberg T. (2022). Stop, children what's that sound? Multi-modal inference through mental simulation. In _Cognitive Science Proceedings_." +publication = "Outa J.\\*, Zhou X. J.\\*, Gweon H., Gerstenberg T. (2022). Stop, children what's that sound? Multi-modal inference through mental simulation. In _Cognitive Science Proceedings_." image_preview = "" abstract = "Human adults can figure out what happened by combining evidence from different sensory modalities, such as vision and sound. How does the ability to integrate multi-modal information develop in early childhood? Inspired by prior computational work and behavioral studies with adults, we examined 3- to 8-year-old children's ability to reason about the physical trajectory of a ball that was dropped into an occluded Plinko box. Children had to infer in which one of three holes the ball was dropped based on visual information (i.e., where the ball landed) and auditory information (i.e., the sounds of the ball colliding with parts of the box). We compare children's responses to the predictions of four computational models. The results suggest that although even the youngest children make systematic judgments rather than randomly guessing, children's ability to integrate visual and auditory evidence continues to develop into late childhood." selected = false diff --git a/content/publication/wu2024whodunnit.md b/content/publication/wu2024whodunnit.md new file mode 100644 index 0000000..5fb8087 --- /dev/null +++ b/content/publication/wu2024whodunnit.md @@ -0,0 +1,33 @@ ++++ +# 0 -> 'Forthcoming', +# 1 -> 'Preprint', +# 2 -> 'Journal', +# 3 -> 'Conference Proceedings', +# 4 -> 'Book chapter', +# 5 -> 'Thesis' + +title = "Whodunnit? Inferring what happened from multimodal evidence" +date = "2024-05-13" +authors = ["S. A Wu\\*","E. Brockbank\\*","H. Cha","J. Fränken","E. Jin","Z. Huang","W. Liu","R. Zhang","J. Wu","T. Gerstenberg"] +publication_types = ["3"] +publication_short = "_Proceedings of the 46th Annual Conference of the Cognitive Science Society_" +publication = "Wu*, S. A., Brockbank*, E., Cha, H., Fränken, J., Jin, E., Huang, Z., Liu, W., Zhang, R., Wu, J., Gerstenberg, T. (2024). Whodunnit? Inferring what happened from multimodal evidence. In _Proceedings of the 46th Annual Conference of the Cognitive Science Society_." +abstract = "Humans are remarkably adept at inferring the causes of events in their environment; doing so often requires incorporating information from multiple sensory modalities. For instance, if a car slows down in front of us, inferences about why they did so are rapidly revised if we also hear sirens in the distance. Here, we investigate the ability to reconstruct others' actions and events from the past by integrating multimodal information. Participants were asked to infer which of two agents performed an action in a household setting given either visual evidence, auditory evidence, or both. We develop a computational model that makes inferences by generating multimodal simulations, and also evaluate our task on a large language model (GPT-4) and a large multimodal model (GPT-4V). We find that humans are relatively accurate overall and perform best when given multimodal evidence. GPT-4 and GPT-4V performance comes close overall, but is very weakly correlated with participants across individual trials. Meanwhile, the simulation model captures the pattern of human responses well. Multimodal event reconstruction represents a challenge for current AI systems, and frameworks that draw on the cognitive processes underlying people's ability to reconstruct events offer a promising avenue forward." +image_preview = "" +selected = false +projects = [] +url_pdf = "papers/wu2024whodunnit.pdf" +url_preprint = "" +url_code = "" +url_dataset = "" +url_slides = "" +url_video = "" +url_poster = "" +url_source = "" +url_custom = [{name = "Github", url = "https://github.com/cicl-stanford/whodunnit_multimodal_inference"}] +math = true +highlight = true +[header] +# image = "publications/wu2024whodunnit.png" +caption = "" ++++ \ No newline at end of file diff --git a/content/publication/zhang2023llm.md b/content/publication/zhang2023llm.md index 112e0a9..d7d5c68 100644 --- a/content/publication/zhang2023llm.md +++ b/content/publication/zhang2023llm.md @@ -8,10 +8,10 @@ title = "You are what you're for: Essentialist categorization in large language models" date = "2023-05-11" -authors = ["S. Zhang","J. S. She","T. Gerstenberg","D. Rose"] +authors = ["S. Zhang\\*","J. S. She\\*","T. Gerstenberg","D. Rose"] publication_types = ["3"] publication_short = "_Proceedings of the 45th Annual Conference of the Cognitive Science Society_" -publication = "Zhang S., She J. S., Gerstenberg T., Rose D. (2023). You are what you're for: Essentialist categorization in large language models. In _Proceedings of the 45th Annual Conference of the Cognitive Science Society_." +publication = "Zhang S.\\*, She J. S.\\*, Gerstenberg T., Rose D. (2023). You are what you're for: Essentialist categorization in large language models. In _Proceedings of the 45th Annual Conference of the Cognitive Science Society_." abstract = "How do essentialist beliefs about categories arise? We hypothesize that such beliefs are transmitted via language. We subject large language models (LLMs) to vignettes from the literature on essentialist categorization and find that they align well with people when the studies manipulated teleological information - information about what something is for. We examine whether in a classic test of essentialist categorization - the transformation task - LLMs prioritize teleological properties over information about what something looks like, or is made of. Experiments 1 and 2 find that telos and what something is made of matter more than appearance. Experiment 3 manipulates all three factors and finds that what something is for matters more than what it's made of. Overall, these studies suggest that language alone may be sufficient to give rise to essentialist beliefs, and that information about what something is for matters more." image_preview = "" selected = false diff --git a/docs/404.html b/docs/404.html index 453cb26..f145e22 100644 --- a/docs/404.html +++ b/docs/404.html @@ -237,6 +237,10 @@

Page not found

Publications

+ + @@ -253,10 +257,6 @@

Publications

  • Do as I explain: Explanations communicate optimal interventions
    • - - diff --git a/docs/bibtex/cic_papers.bib b/docs/bibtex/cic_papers.bib index c883a71..41ea574 100644 --- a/docs/bibtex/cic_papers.bib +++ b/docs/bibtex/cic_papers.bib @@ -1,13 +1,23 @@ %% This BibTeX bibliography file was created using BibDesk. %% https://bibdesk.sourceforge.io/ -%% Created for Tobias Gerstenberg at 2024-05-12 17:48:57 +0200 +%% Created for Tobias Gerstenberg at 2024-05-13 09:25:20 +0200 %% Saved with string encoding Unicode (UTF-8) +@inproceedings{wu2024whodunnit, + abstract = {Humans are remarkably adept at inferring the causes of events in their environment; doing so often requires incorporating information from multiple sensory modalities. For instance, if a car slows down in front of us, inferences about why they did so are rapidly revised if we also hear sirens in the distance. Here, we investigate the ability to reconstruct others' actions and events from the past by integrating multimodal information. Participants were asked to infer which of two agents performed an action in a household setting given either visual evidence, auditory evidence, or both. We develop a computational model that makes inferences by generating multimodal simulations, and also evaluate our task on a large language model (GPT-4) and a large multimodal model (GPT-4V). We find that humans are relatively accurate overall and perform best when given multimodal evidence. GPT-4 and GPT-4V performance comes close overall, but is very weakly correlated with participants across individual trials. Meanwhile, the simulation model captures the pattern of human responses well. Multimodal event reconstruction represents a challenge for current AI systems, and frameworks that draw on the cognitive processes underlying people's ability to reconstruct events offer a promising avenue forward.}, + author = {Wu, Sarah A and Brockbank, Erik and Cha, Hannah and Fr{\"a}nken, Jan-Philipp and Jin, Emily and Huang, Zhuoyi and Liu, Weiyu and Zhang, Ruohan and Wu, Jiajun and Gerstenberg, Tobias}, + booktitle = {{Proceedings of the 46th Annual Conference of the Cognitive Science Society}}, + date-added = {2024-05-13 09:25:14 +0200}, + date-modified = {2024-05-13 09:25:14 +0200}, + editor = {Larissa K Samuelson and Stefan Frank and Mariya Toneva and Allyson Mackey and Eliot Hazeltine}, + title = {{Whodunnit? Inferring what happened from multimodal evidence}}, + year = {2024}} + @inproceedings{tsirtsis2024sequential, abstract = {When a human and an AI agent collaborate to complete a task and something goes wrong, who is responsible? Prior work has developed theories to describe how people assign responsibility to individuals in teams. However, there has been little work studying the cognitive processes that underlie responsibility judgments in human-AI collaborations, especially for tasks comprising a sequence of interdependent actions. In this work, we take a step towards filling this gap. Using semiautonomous driving as a paradigm, we develop an environment that simulates stylized cases of human-AI collaboration using a generative model of agent behavior. We propose a model of responsibility that considers how unexpected an agent's action was, and what would have happened had they acted differently. We test the model's predictions empirically and find that in addition to action expectations and counterfactual considerations, participants' responsibility judgments are also affected by how much each agent actually contributed to the outcome.}, author = {Tsirtsis, Stratis and Gomez-Rodriguez, Manuel and Gerstenberg, Tobias}, diff --git a/docs/index.html b/docs/index.html index 13b8c47..47dd1dc 100644 --- a/docs/index.html +++ b/docs/index.html @@ -110,7 +110,7 @@ - + diff --git a/docs/index.xml b/docs/index.xml index 5f0102f..94ade24 100644 --- a/docs/index.xml +++ b/docs/index.xml @@ -6,9 +6,18 @@ Hugo -- gohugo.io en-us © 2024 Tobias Gerstenberg - Sun, 12 May 2024 00:00:00 +0000 + Mon, 13 May 2024 00:00:00 +0000 + + Whodunnit? Inferring what happened from multimodal evidence + https://cicl.stanford.edu/publication/wu2024whodunnit/ + Mon, 13 May 2024 00:00:00 +0000 + + https://cicl.stanford.edu/publication/wu2024whodunnit/ + + + Towards a computational model of responsibility judgments in sequential human-AI collaboration https://cicl.stanford.edu/publication/tsirtsis2024sequential/ @@ -135,14 +144,5 @@ - - MoCa: Measuring human-language model alignment on causal and moral judgment tasks - https://cicl.stanford.edu/publication/nie2023moca/ - Fri, 20 Oct 2023 00:00:00 +0000 - - https://cicl.stanford.edu/publication/nie2023moca/ - - - diff --git a/docs/member/tobias_gerstenberg/index.html b/docs/member/tobias_gerstenberg/index.html index 1c62745..57c614a 100644 --- a/docs/member/tobias_gerstenberg/index.html +++ b/docs/member/tobias_gerstenberg/index.html @@ -356,6 +356,49 @@

    Publications

    + + + (2024). + + Whodunnit? Inferring what happened from multimodal evidence. + Proceedings of the 46th Annual Conference of the Cognitive Science Society. + + + + +

    + + + + + + + PDF + + + + + + + + + + + + + + + + + Github + + + +

    + +
    +
    @@ -1463,7 +1506,7 @@

    Publications

    + S. Zhang*, J. S. She*, T. Gerstenberg, D. Rose (2023). @@ -1514,7 +1557,7 @@

    Publications

    + A. Cao*, A. Geiger*, E. Kreiss*, T. Icard, T. Gerstenberg (2023). @@ -2002,7 +2045,7 @@

    Publications

    + J. Outa*, X. Jia Zhou*, H. Gweon, T. Gerstenberg (2022). @@ -3781,7 +3824,7 @@

    Publications

    + S. J. Gershman, T. Gerstenberg, C. L. Baker, F. Cushman (2016). @@ -4759,7 +4802,7 @@

    Publications

    + J. McCoy*, T. D. Ullman*, A. Stuhlmüller, T. Gerstenberg, J. B. Tenenbaum (2012). diff --git a/docs/papers/wu2024whodunnit.pdf b/docs/papers/wu2024whodunnit.pdf new file mode 100644 index 0000000..bf2aafe Binary files /dev/null and b/docs/papers/wu2024whodunnit.pdf differ diff --git a/docs/publication/cao2023semantics/index.html b/docs/publication/cao2023semantics/index.html index 07efbe0..a37315b 100644 --- a/docs/publication/cao2023semantics/index.html +++ b/docs/publication/cao2023semantics/index.html @@ -239,7 +239,7 @@

    A Semantics for Causing, Enabling, and Preventing Verbs Using Structural Causal Models

    @@ -283,7 +283,7 @@

    Abstract

    Publication
    -
    Cao A., Geiger A., Kreiss E., Icard T., Gerstenberg T. (2023). A Semantics for Causing, Enabling, and Preventing Verbs Using Structural Causal Models. In Proceedings of the 45th Annual Conference of the Cognitive Science Society.
    +
    Cao A.*, Geiger A.*, Kreiss E.*, Icard T., Gerstenberg T. (2023). A Semantics for Causing, Enabling, and Preventing Verbs Using Structural Causal Models. In Proceedings of the 45th Annual Conference of the Cognitive Science Society.
    diff --git a/docs/publication/gershman2016goals/index.html b/docs/publication/gershman2016goals/index.html index dd19cf9..bb11f6c 100644 --- a/docs/publication/gershman2016goals/index.html +++ b/docs/publication/gershman2016goals/index.html @@ -239,7 +239,7 @@

    Plans, habits, and theory of mind

    @@ -283,7 +283,7 @@

    Abstract

    Publication
    -
    Gershman, S. J., Gerstenberg, T., Baker, C. L., & Cushman, F. (2016). Plans, habits, and theory of mind. PLoS ONE, 11(9), e0162246.
    +
    Gershman, S. J.*, Gerstenberg, T.*, Baker, C. L., & Cushman, F. (2016). Plans, habits, and theory of mind. PLoS ONE, 11(9), e0162246.
    diff --git a/docs/publication/index.html b/docs/publication/index.html index 2359687..f6e35e6 100644 --- a/docs/publication/index.html +++ b/docs/publication/index.html @@ -1545,6 +1545,19 @@

    Publications

    + + + + + + + + + + + + + @@ -1681,6 +1694,61 @@

    Publications

    +
    + +
    + + + (2024). + + Whodunnit? Inferring what happened from multimodal evidence. + Proceedings of the 46th Annual Conference of the Cognitive Science Society. + + + + +

    + + + + + + + PDF + + + + + + + + + + + + + + + + + Github + + + +

    + +
    + + +
    + + + + + + +
    @@ -3079,7 +3147,7 @@

    Publications

    + S. Zhang*, J. S. She*, T. Gerstenberg, D. Rose (2023). @@ -3142,7 +3210,7 @@

    Publications

    + A. Cao*, A. Geiger*, E. Kreiss*, T. Icard, T. Gerstenberg (2023). @@ -3750,7 +3818,7 @@

    Publications

    + J. Outa*, X. Jia Zhou*, H. Gweon, T. Gerstenberg (2022). @@ -5973,7 +6041,7 @@

    Publications

    + S. J. Gershman, T. Gerstenberg, C. L. Baker, F. Cushman (2016). @@ -7215,7 +7283,7 @@

    Publications

    + J. McCoy*, T. D. Ullman*, A. Stuhlmüller, T. Gerstenberg, J. B. Tenenbaum (2012). diff --git a/docs/publication/index.xml b/docs/publication/index.xml index 7b3091b..dd8041b 100644 --- a/docs/publication/index.xml +++ b/docs/publication/index.xml @@ -12,6 +12,15 @@ + + Whodunnit? Inferring what happened from multimodal evidence + https://cicl.stanford.edu/publication/wu2024whodunnit/ + Mon, 13 May 2024 00:00:00 +0000 + + https://cicl.stanford.edu/publication/wu2024whodunnit/ + + + Towards a computational model of responsibility judgments in sequential human-AI collaboration https://cicl.stanford.edu/publication/tsirtsis2024sequential/ diff --git a/docs/publication/mccoy2012probabilistic/index.html b/docs/publication/mccoy2012probabilistic/index.html index 70fbece..6ca5c40 100644 --- a/docs/publication/mccoy2012probabilistic/index.html +++ b/docs/publication/mccoy2012probabilistic/index.html @@ -239,7 +239,7 @@

    Why blame Bob? Probabilistic generative models, counterfactual reasoning, and blame attribution

    @@ -283,7 +283,7 @@

    Abstract

    Publication
    -
    McCoy, J., Ullman, T., Stuhlmüller, A., Gerstenberg, T., & Tenenbaum, J. B. (2012). Why blame Bob? Probabilistic generative models, counterfactual reasoning, and blame attribution. In Proceedings of the 34th Annual Conference of the Cognitive Science Society (pp. 1996-2001). Austin, TX: Cognitive Science Society.
    +
    McCoy, J.*, Ullman, T.*, Stuhlmüller, A., Gerstenberg, T., & Tenenbaum, J. B. (2012). Why blame Bob? Probabilistic generative models, counterfactual reasoning, and blame attribution. In Proceedings of the 34th Annual Conference of the Cognitive Science Society (pp. 1996-2001). Austin, TX: Cognitive Science Society.
    diff --git a/docs/publication/outa2022stop/index.html b/docs/publication/outa2022stop/index.html index a555eea..ac9dc5c 100644 --- a/docs/publication/outa2022stop/index.html +++ b/docs/publication/outa2022stop/index.html @@ -248,7 +248,7 @@

    Stop, children what's that sound? Multi-modal inference through mental simulation

    @@ -292,7 +292,7 @@

    Abstract

    Publication
    -
    Outa* J., Zhou* X. J., Gweon H., Gerstenberg T. (2022). Stop, children what’s that sound? Multi-modal inference through mental simulation. In Cognitive Science Proceedings.
    +
    Outa J.*, Zhou X. J.*, Gweon H., Gerstenberg T. (2022). Stop, children what’s that sound? Multi-modal inference through mental simulation. In Cognitive Science Proceedings.
    diff --git a/docs/publication/wu2024whodunnit/index.html b/docs/publication/wu2024whodunnit/index.html new file mode 100644 index 0000000..dbab959 --- /dev/null +++ b/docs/publication/wu2024whodunnit/index.html @@ -0,0 +1,492 @@ + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + Whodunnit? Inferring what happened from multimodal evidence | Causality in Cognition Lab + + + + + + +
    + + + + +
    +
    +

    Whodunnit? Inferring what happened from multimodal evidence

    + + + + + +
    + + + + +

    Abstract

    +

    Humans are remarkably adept at inferring the causes of events in their environment; doing so often requires incorporating information from multiple sensory modalities. For instance, if a car slows down in front of us, inferences about why they did so are rapidly revised if we also hear sirens in the distance. Here, we investigate the ability to reconstruct others’ actions and events from the past by integrating multimodal information. Participants were asked to infer which of two agents performed an action in a household setting given either visual evidence, auditory evidence, or both. We develop a computational model that makes inferences by generating multimodal simulations, and also evaluate our task on a large language model (GPT-4) and a large multimodal model (GPT-4V). We find that humans are relatively accurate overall and perform best when given multimodal evidence. GPT-4 and GPT-4V performance comes close overall, but is very weakly correlated with participants across individual trials. Meanwhile, the simulation model captures the pattern of human responses well. Multimodal event reconstruction represents a challenge for current AI systems, and frameworks that draw on the cognitive processes underlying people’s ability to reconstruct events offer a promising avenue forward.

    + + + + + + +
    +
    +
    +
    +
    Type
    +
    + + + Conference Proceedings + + +
    +
    +
    +
    +
    +
    + + +
    +
    +
    +
    +
    Publication
    +
    Wu, S. A., Brockbank, E., Cha, H., Fränken, J., Jin, E., Huang, Z., Liu, W., Zhang, R., Wu, J., Gerstenberg, T. (2024). Whodunnit? Inferring what happened from multimodal evidence. In Proceedings of the 46th Annual Conference of the Cognitive Science Society.
    +
    +
    +
    +
    +
    + +
    +
    +
    +
    +
    Date
    + +
    +
    +
    +
    +
    + + +
    +
    +
    +
    +
    Links
    +
    + + + + + + + PDF + + + + + + + + + + + + + + + + + Github + + + +
    +
    +
    +
    +
    +
    + + +
    + +
    + +

    << Back to list of publications

     +
    + +
    + + + + + +
    + + +
    + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + diff --git a/docs/publication/zhang2023llm/index.html b/docs/publication/zhang2023llm/index.html index b664f14..03df3e7 100644 --- a/docs/publication/zhang2023llm/index.html +++ b/docs/publication/zhang2023llm/index.html @@ -239,7 +239,7 @@

    You are what you're for: Essentialist categorization in large language models

    @@ -283,7 +283,7 @@

    Abstract

    Publication
    -
    Zhang S., She J. S., Gerstenberg T., Rose D. (2023). You are what you’re for: Essentialist categorization in large language models. In Proceedings of the 45th Annual Conference of the Cognitive Science Society.
    +
    Zhang S.*, She J. S.*, Gerstenberg T., Rose D. (2023). You are what you’re for: Essentialist categorization in large language models. In Proceedings of the 45th Annual Conference of the Cognitive Science Society.
    diff --git a/docs/publication_types/3/index.html b/docs/publication_types/3/index.html index 6925112..5c97d1e 100644 --- a/docs/publication_types/3/index.html +++ b/docs/publication_types/3/index.html @@ -111,7 +111,7 @@ - + @@ -238,6 +238,15 @@

    3

    +
    +

    Whodunnit? Inferring what happened from multimodal evidence

    +
    + + Humans are remarkably adept at inferring the causes of events in their environment; doing so often requires incorporating information from multiple sensory modalities. For instance, if a car slows down in front of us, inferences about why they did so … + +
    +
    + -
    -

    MoCa: Measuring human-language model alignment on causal and moral judgment tasks

    -
    - - Human commonsense understanding of the physical and social world is organized around intuitive theories. These theories support making causal and moral judgments. When something bad happens, we naturally ask: who did what, and why? A rich literature … - -
    -
    -