INSTINCT: Investigating Novel Statistical Techniques to Identify Neurophysiological Correlates of Trustworthiness
Whom do you trust? Why do you trust them? How do you know whether to trust someone you’ve just met? The answers to these questions are essential in everyday interactions but particularly so in the Intelligence Community, where knowing whom to trust is often vital. The Intelligence Advanced Research Project Activity (IARPA) TRUST program seeks ways to detect one’s own neural, psychological, physiological, and behavioral signals that reflect a partner’s trustworthiness. The goal of this Challenge is to develop an algorithm that identifies and extracts such signals from data recorded while volunteers engaged in various types of trust activities. Cross-disciplinary teaming is encouraged in order to bring together expertise from diverse fields (such as neurophysiology and data analytics) to solve this complex problem.
This is a Reduction-to-Practice Challenge that requires written documentation and delivery of source code implementing an algorithm that solves the problem. This is also a Prodigy Challenge and a real-time online scoring utility and leaderboard will be available to track Solver algorithm performance.
There will be up to 3 awards: $25,000 for first place, $15,000 for second place, and $10,000 for third place. Awards will be based on Seeker’s determination of solution performance using a reserved independent validation set.
Trust plays a fundamental role in many human relationships, organizations, and behaviors. Knowing whom to trust is especially vital for many Intelligence Community missions and organizations. Starting from the premise that people – often non-consciously – generate neural, psychological, physiological, and behavioral signals in response to others’ trustworthiness cues, the IARPA TRUST (Tools for Recognizing Useful Signals of Trustworthiness) program seeks to develop capabilities to detect, measure, and validate one’s own “useful” signals in order to more accurately assess another’s trustworthiness in a particular context. Improving the accuracy of judgments about whom can be trusted and under what conditions could have profound implications for not just the Intelligence Community, but society in general.
In a series of recent research studies funded by IARPA, voluntary participants interacted with other volunteers while undertaking a number of tasks that required each of them to assess the other’s trustworthiness. In turn, each participant had to decide whether they would act in a trustworthy fashion towards their partner. Importantly, both participants could gain or lose stakes based on the combined consequences of each person’s willingness to trust and each person’s willingness and ability to keep specific promises made to the other. Neural, psychological, and physiological data were collected in parallel with these tasks, with participants’ behavior serving as ground truth (i.e., partners did or did not keep their promises). The Air Force Research Laboratory (AFRL) has conducted preliminary analyses of these data, and now joins IARPA in inviting Solvers to explore the data in greater depth.
The Seekers (IARPA and AFRL) ask Challenge Solvers to create innovative algorithms and analyses that use data from one participant in order to predict whether that participant’s task partners will act in a trustworthy manner. Solvers will be provided with sample data including labeled trustworthy and non-trustworthy behaviors. They will be asked to develop techniques and models based on these data, and then submit predictions for an unlabeled test set.
Data are made available with this Challenge solely for the purposes of advancing the scientific study of human trust. Any efforts to identify participants, sell the data, or use the data without permission are strictly forbidden. Misuse of data is a violation of terms and may result in disqualification.
This is a Reduction-to-Practice Challenge that requires written documentation and delivery of source code implementing an algorithm that solves the problem.
The award is contingent upon evaluation and validation of the submitted Solutions by the Seeker. During the evaluation period, the Seeker will validate top-scoring submissions using additional data similar to the training data provided in the Challenge.
To receive an award, Solvers will be required to grant the United States Government certain rights, detailed in the Challenge Specific Agreement (CSA), for United States Government purposes. Commercial rights will remain the property of the authors/inventors. The United States Government’s rights in the Solution IP will not limit the rights of Solvers to use, release, perform display, disclose or publish their submission. However, the Solution IP will remain subject to any other applicable restrictions (e.g., Export Control).
Federal entities or Federal employees acting within the scope of their employment are NOT eligible to participate in this Challenge. Each Solver is eligible to receive at most one award.
NOTE: Awards for this Challenge may only be paid to Solvers who are citizens or permanent residents of the United States, or private entities that are incorporated in and maintaining a primary place of business in the United States. While non-U.S. citizens or non-permanent residents may participate as members of a team or organization led by a U.S. citizen, permanent resident, or private entity, only U.S. entities, citizens, and permanent residents are eligible to receive an award for this Challenge. Please do not submit if you do not meet this requirement. If you have a question about eligibility, please use your Project Room to ask.
By making a submission to this Challenge, Solver(s) are providing written consent to the Seeker to publicly disclose their identity if their submission is chosen for an award and they choose to accept the award. Disclosure will not occur until the Challenge has been fully completed, that is, all submissions have been evaluated, rejected or accepted, and any awards have been transferred to the Solver(s).
About the Seeker:
This Challenge is sponsored by the Intelligence Advanced Research Projects Activity (IARPA), in partnership with the Air Force Research Laboratory (AFRL). IARPA invests in high-risk, high-payoff research programs that have the potential to provide the United States with an overwhelming intelligence advantage over future adversaries. AFRL is the Air Force's only organization wholly dedicated to leading the discovery, development, and integration of warfighting technologies for our air, space and cyberspace forces. Opinions, interpretations, recommendations and conclusions are those of the authors and are not necessarily endorsed by the United States Government.
What is an RTP Challenge?
An InnoCentive RTP (Reduction to Practice) Challenge is a prototype that proves an idea, and is similar to an InnoCentive Theoretical Challenge in its high level of detail. However, an RTP requires the Solver to submit a validated solution, either in the form of original data or a physical sample. Also the Seeker is allowed to test the proposed solution. For details about treatment of IP rights, please see the Challenge-Specific Agreement.