Phil 2.19.18

7:30 – 4:30 ASRC MKT

  • Back to BIC.
    • BIC_102 (page 102)
    • BIC107 (pg 107)
    • BIC107b (pg 107)
    • Sociality: Coordinating bodies, minds and groups
      • Human interaction, as opposed to aggregation, occurs in face-to-face groups. “Sociality theory” proposes that such groups have a nested, hierarchical structure, consisting of a few basic variations, or “core configurations.” These function in the coordination of human behavior, and are repeatedly assembled, generation to generation, in human ontogeny, and in daily life. If face-to-face groups are “the mind’s natural environment,” then we should expect human mental systems to correlate with core configurations. Features of groups that recur across generations could provide a descriptive paradigm for testable and non-intuitive evolutionary hypotheses about social and cognitive processes. This target article sketches three major topics in sociality theory, roughly corresponding to the interests of biologists, psychologists, and social scientists. These are (1) a multiple levels-of-selection view of Darwinism, part group selectionism, part developmental systems theory; (2) structural and psychological features of repeatedly assembled, concretely situated face-to-face coordination; and (3) superordinate, “unsituated” coordination at the level of large-scale societies. Sociality theory predicts a tension, perhaps unresolvable, between the social construction of knowledge, which facilitates coordination within groups, and the negotiation of the habitat, which requires some correspondence with contingencies in specific situations. This tension is relevant to ongoing debates about scientific realism, constructivism, and relativism in the philosophy and sociology of knowledge.
        • These definitions seem to span atomic (mother/child, etc), small group (situated, environmental), and societal (unsituated, normative)
      • Coordination occurs to the extent that knowledge and practice domains overlap or are complementary. I suggest that values serve as a medium. Humans live in a value-saturated environment; values are known from interactions with people, natural objects, and artifacts
        • Dimension reduction
  •  I’m starting to think that agents as gradient descent machines within networks is something to look for:
    • Individual Strategy Update and Emergence of Cooperation in Social Networks
      • In this article, we critically study whether social networks can explain the emergence of cooperative behavior. We carry out an extensive simulation program in which we study the most representative social dilemmas. For the Prisoner’s Dilemma, it turns out that the emergence of cooperation is dependent on the microdynamics. On the other hand, network clustering mostly facilitates global cooperation in the Stag Hunt game, whereas degree heterogeneity promotes cooperation in Snowdrift dilemmas. Thus, social networks do not promote cooperation in general, because the macro-outcome is not robust under change of dynamics. Therefore, having specific applications of interest in mind is crucial to include the appropriate microdetails in a good model.
    • Alex Peysakhovich and Adam Lerer
      • Prosocial learning agents solve generalized Stag Hunts better than selfish ones
        • Deep reinforcement learning has become an important paradigm for constructing agents that can enter complex multi-agent situations and improve their policies through experience. One commonly used technique is reactive training – applying standard RL methods while treating other agents as a part of the learner’s environment. It is known that in general-sum games reactive training can lead groups of agents to converge to inefficient outcomes. We focus on one such class of environments: Stag Hunt games. Here agents either choose a risky cooperative policy (which leads to high payoffs if both choose it but low payoffs to an agent who attempts it alone) or a safe one (which leads to a safe payoff no matter what). We ask how we can change the learning rule of a single agent to improve its outcomes in Stag Hunts that include other reactive learners. We extend existing work on reward-shaping in multi-agent reinforcement learning and show that that making a single agent prosocial, that is, making them care about the rewards of their partners can increase the probability that groups converge to good outcomes. Thus, even if we control a single agent in a group making that agent prosocial can increase our agent’s long-run payoff. We show experimentally that this result carries over to a variety of more complex environments with Stag Hunt-like dynamics including ones where agents must learn from raw input pixels.
      • The Good, the Bad, and the Unflinchingly Selfish: Cooperative Decision-Making Can Be Predicted with High Accuracy Using Only Three Behavioral Types
        • The human willingness to pay costs to benefit anonymous others is often explained by social preferences: rather than only valuing their own material payoff, people also care in some fashion about the outcomes of others. But how successful is this concept of outcome-based social preferences for actually predicting out-of-sample behavior? We investigate this question by having 1067 human subjects each make 20 cooperation decisions, and using machine learning to predict their last 5 choices based on their first 15. We find that decisions can be predicted with high accuracy by models that include outcome-based features and allow for heterogeneity across individuals in baseline cooperativeness and the weights placed on the outcome-based features (AUC=0.89). It is not necessary, however, to have a fully heterogeneous model — excellent predictive power (AUC=0.88) is achieved by a model that allows three different sets of baseline cooperativeness and feature weights (i.e. three behavioral types), defined based on the participant’s cooperation frequency in the 15 training trials: those who cooperated at least half the time, those who cooperated less than half the time, and those who never cooperated. Finally, we provide evidence that this inclination to cooperate cannot be well proxied by other personality/morality survey measures or demographics, and thus is a natural kind (or “cooperative phenotype”)
        • “least”, “intermediate” and “most” cooperative. Doesn’t give percentages, though it says that 17.8% were cooperative?

         

  • Talk Susan Gregurick (susan.gregurick@nih.gov)
    • All of Us research program
    • Opiod epidemic – trajectory modeling?
    • PZM21 computational drug
    • Develop advanced software and tools. Specialized generalizable and accessible tools for biomedicing (finding stream). Includes mobile, data indexing, etc.
    • NIH Data Fellows? Postdocs to senior industry
    • T32 funding? Mike Summers at UMBC
    • ncbi-hackathons.github.io (look for data?
    • Primary supporter for machine learning is NIMH (imaging), then NIGNS, and NCI Team science (Multi-PI) is a developing thing
    • $400m in computing enabled interactions (human in the loop decision tools. Research Browser?
    • Big Data to Knowledge Initiative (BD2K) datascience.nih.gov/bd2k
    • Interagency Modeling and Analysis Group (IMAG) imagewiki,nibib.nih.gov
    • funding: bisti.nih.gov
    • NIH RePorter projectreporter.nih.gov Check out matchmaker. What’s the ranking algorithm?
    • NIDDK predictive analytics for budgeting <- A2P-ish?
    • Most of thi srequires preliminary data and papers to be considered for funding. There is one opportunity for getting funding to get preliminary data. Need to get more specific infor here.
    • Each SRO normalizes grade as a percentile, not the score, since some places inflate, and others are hard.
    • Richard Aargon at NIGMS
    • Office of behavioral and social science – NIH center Francis Collins. Also agent-based simulation
    • Really wants a Research Browser to go through proposals
  • Fika – study design
    • IRB – you can email and chat with the board if you have a tricky study

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.