Call for 2024 CCN GAC Proposals

What is a Generative Adversarial Collaboration (GAC)? A brief introduction.

What are the most exciting challenges in cognitive computational neuroscience and how might we solve them together? The goal of a CCN GAC is to make concrete plans to address important debates in our field -- and then to begin implementing these plans.

A GAC is a collaboration between scientists to come together and formulate a plan to resolve opposing viewpoints via a jointly designed research agenda. The goal is to provide a forum for bridging scientific perspectives that might otherwise be challenging to reconcile. Ultimately, we aim for such activities to lead to better scientific discourse and accelerate scientific progress.

What problem does a GAC address that is not addressed by "normal" symposia or collaborations?

Most science proceeds with competing viewpoints investigated and published in parallel, following separate lines of inquiry, and by separate groups. Scientists converge on points of contention or agreement—and discover potential collaborators—through reading these publications, interacting informally at conferences, and other largely one-way or asynchronous methods of communication. In the case of competing theories, this process often leads to further series of parallel, separate experiments being performed, where one group claims the results support their view, while another group argues it wasn’t the right experiment or a key control condition is missing.

The CCN GAC project aims to foster direct debates and interactions between theories, thereby accelerating scientific progress and leading to novel collaborations. 

The three elements of a GAC: collaboration, discourse (“adversarial” interaction between theories), and generativity.

Collaboration: Collaboration and community are front and center for a GAC. The goal of everyone involved is to make progress as a field. GACs bring together scientists from different communities who may not have interacted scientifically before. A GAC provides a mechanism to strengthen bridges between communities by developing a shared language and mutual understanding. It is not uncommon for GAC members to discover more areas of agreement than initially thought. 

Discourse, i.e. “adversarial” interaction between theories: Adjudicating among competing theories is core to the scientific method. In a GAC, the adversarial component comes from competition between theories, and the generative and collaborative components come from constructive interactions among the people involved. By “adversarial collaboration” we mean “a good-faith effort to conduct debates by carrying out joint research” (Kahneman, 2003, p.729; Latham et al., 1988).

Generativity: GACs generate many benefits for the organizers, the CCN community, and our field overall. Foremost, a shared language is created facilitating the exchange between communities. The adversarial interaction between theories will likely also generate novel questions, hypotheses, and theoretical positions, and lead to new experiments. A tangible outcome of a GAC is a joint position paper.

>> Check out the 15 GACs from previous years

>> You can read more about the GAC project in the preprint
"Generative Adversarial Collaborations: A practical guide for conference organizers and participating scientists"

How does the GAC process work in detail?

Each selected GAC will be carefully facilitated with the support of CCN over the coming months, including explicit feedback and interactive participation from the CCN community, and then kicked off with a virtual debate and discussion event—all with the targeted goal of developing proposals to arbitrate the competing hypotheses or models. The GAC team will then write a paper together laying out their plan for resolving the controversy through experiment, simulation, or theory. These papers are expected to clarify the controversy or open question and to propose concrete steps—experimental, theoretical or a combination—as a challenge to the larger community to engage. CCN is coordinating with the Neurons, Behavior, Data analysis, and Theory journal to publish the GAC papers as a special issue along with accompanying peer-reviewed commentary pieces from community members inspired by the Behavioral and Brain Sciences model. We will then invite all the GACs to present their progress with the plan they proposed in the paper at the following CCN. With this ambitious programme we aim to significantly propel scientific progress on foundational questions and help spark ideas for future research projects and grant proposals resulting from the CCN GACs (see also Ellemers et al., 2020, Lakens 2020). 

GAC workshop formation

Phase 1. First, you will need to form a core team, consisting of at least 2 members with opposing viewpoints (or willing to espouse competing views/theories for the purpose of the GAC) who are willing to both debate and work collaboratively together on the next steps. The core team can also contain several other members of your choosing whom you feel can and will substantively contribute to developing ideas and exploring avenues for possible resolution of the controversy. This core team will together write and submit the letter of intent, followed by the proposal for the CCN GAC workshop and subsequent rapid invited special issue publication. A small number of proposals will be selected for participation in the  CCN GAC kickoff workshops, to take place in person at the CCN annual meeting and subsequent special issue (see below for timeline).

Phase 2. The CCN program comittee will select which GAC will participate in this cycle’s CCN GAC workshops. The structure of the proposed GAC workshop at CCN will play an important role during the selection process. All selected proposals will be published online, and we will solicit comments, reactions, and feedback from the whole community through an open discussion and feedback process. Community members may choose to submit comments, critiques, and reactions to as many of the GAC proposals as they wish, although we encourage feedback to be substantive and constructive: “Cool idea!” is encouraging, but the goal of this feedback is to strengthen the GAC, and potentially allow the core GAC team to identify select community members who could join the GAC and help drive it toward fruitful outcome. Relevant theoretical background, conceptual observations, and/or concrete ideas of experimental paradigms or approaches which further understanding or potential resolution of the competing views are highly encouraged. 

Phase 3. All community comments and reactions will be shared with the selected GAC core teams to encourage them to bring on board those who have demonstrated commitment and potentially valuable contributions. The core GAC teams will then integrate this community feedback into their GAC, and potentially invite community members to join the GAC whom they feel can substantively contribute to the project.

Phase 4. The (potentially expanded) GAC team will organize and then run a GAC workshop that we encourage to include clear presentation of the controversy or open question, debate about points of uncertainty or contention, and explore potential avenues to resolve or move forward at the CCN GAC workshops as part of the CCN annual meeting (see timeline below).

Phase 5: After the workshops, the GAC team is expected to collectively and collaboratively write a position paper that outlines theoretical and/or experimental proposals for submission to a special issue and to continue working on their project throughout the following year culminating in a presentation of their accomplishments at the following CCN annual meeting. The position paper is not meant to necessarily provide a resolution to any debate, but rather to document the outcome of the workshop and any insights, vocabulary alignments, or new lines of thinking that may have emerged. It also need not be a “consensus” paper, and can present separate viewpoints supported by different authors in a flexible format. If GAC teams have questions about acceptable formats for these position pieces, they can refer to previous papers or reach out to the CCN GAC team (gac@ccneuro.org).

In-person GAC workshop

CCN GAC workshops are expected to be highly interactive, starting with an overview of the question and a debate between two or more proponents of competing hypotheses or theories and then moving to commentaries and reactions from the rest of the GAC team. Importantly, we expect these CCN GACs not to be a series of isolated talks, but instead an integrated presentation of competing viewpoints and open questions focused on moving the scientific process forward. Key components of the debate and discussion should include focus on identifying conflicting experimental results and brainstorming specific experimental approaches and potential empirical (or theoretical) outcomes that might arbitrate the competing theories. 

CCN GAC workshops (debates + commentary) will be recorded and broadcast to an audience with a capacity of thousands (CCN youtube channel); interaction from the audience is highly encouraged during the live event. We also encourage the use of breakout sessions with the audience, polling or live surveys, soliciting open comments, and other ways of actively engaging the CCN community for the benefit of the GAC topic. 

All previous GAC proposals, recordings, and position papers, can be found here. Among many great GACs, there are two examples in particular, that we would like to highlight as successful examples:

• An excellent example of the scientific and educational value of pitting competing theories against each other is the 2020 GAC on Competing Theories of Probabilistic Computations in the Brain. 

• A wonderful example of how to make a GAC workshop accessible, interactive, and entertaining for the CCN audience is the 2022 GAC To what extent does the brain simulate the external world? 

Note, that your plan for an engaging workshop for the CCN community is a crucial element considered during the selection of proposals by the programme committee

GAC special issue manuscript

After the initial debate and workshop, the GAC team is expected to develop and submit the invited submission manuscript to the publication venue arranged by CCN. 

Workshop attendees & special issue commentaries: Even though they are not part of the GAC core team selected to organize the workshop, audience members/attendees of the CCN GAC workshops and other community members are all invited to contribute commentary pieces, to be peer-reviewed for publication as accompanying pieces to the main GAC position paper in a special issue (journal arranged by CCN). Thus, through the GAC workshops, main position paper, and accompanying commentaries, the broader community can come to serve as public discussants of (or even advisors to, in some cases) the main project even if they do not become involved as core GAC members.

Target GAC outcomes - summary

In sum, selected GAC proposals will participate in the following:

1. Co-organize and attend a kickoff CCN GAC workshop (in person) during the CCN annual meeting.

2. Write a collaborative paper based on the CCN GAC, to be submitted to a special issue of NBDT. This position paper can identify the concrete goals (theoretical or experimental) of the GAC over the next several years, discuss new issues arising from the workshop, or otherwise lay out a path forward as identified by the GAC organizers. These papers may be accompanied by commentary pieces (especially from workshop audience members) submitted for peer review to the same journal, to be linked with the collaborative manuscript.

3. Attend and present progress at the following CCN annual meeting.

Commitment from CCN

For the CCN GACs and future related products, CCN will:

• Provide a platform for the curated discussion of GAC topics and assist with logistical organization of the workshops

• Foster and support the development of the collaborations (help identify missing members, help with experimental/theory interactions, guidance on developing successful distributed collaborations, etc.) 

• Solicit and organize feedback on the GAC proposals from the community

• Coordinate with the NBDT journal for the CCN GAC publications and community involvement through accompanying commentary pieces

• Organize in-person presentations and discussion of the CCN GAC outcomes at CCN annual meetings

Commitments from the GAC team members

The core members of a GACs team commit to:

• Submitting a proposal and Incorporating feedback from the community

• Potentially welcoming new CCN community members to the GAC based on their written commentary 

• Running an in-person kickoff workshop at the annual CCN meeting, inclusive of both founding core GAC members and those new members who joined through the community feedback process

• Writing the position paper to be submitted to a curated special issue of NBDT

• Attending and presenting progress at the following CCN annual meeting

Application process

Please refer to ccneuro.org for open calls and how to submit your proposal. We also strongly encourage you to read the paper Generative Adversarial Collaborations: A practical guide for conference organizers and participating scientists, summarizing the experience and lessons learned both from a perspective of participating teams and the organizers.

Selection process

A small number of CCN GAC proposals will be accepted for the CCN GAC workshops and following development during the year. The selection process will involve the CCN Program Committee and Review Committee evaluating the proposals for (1) adherence to the above criteria and (2) diversity and breadth of target audience, (3) the anticipated attractiveness of the in-person GAC workshop for the CCN audience, and will incorporate feedback solicited from the CCN community in the process.

Timeline targets (subject to revision)

Please refer to ccneuro.org for up-to-date information!

Leading up to the CCN annual meeting:

• February - Call for proposals

• March - Deadline for proposals

• Early May - Decision over GAC selection. Start of the community feedback phase.

• July  - Final list of GAC discussants, proposed workshop schedule

• Aug - GAC kickoff workshops @ CCN annual meeting!

Following CCN annual meeting:

• Work on and submit a manuscript (GAC team) to a special issue of NBDT.

• Aug/Sept of the following year - GACs present their progress at CCN annual meeting.

FAQs

I have an idea for a symposium, that does not involve competing viewpoints. Does a GAC have to be centered around competing viewpoints and theories (i.e., be “adversarial”)?

Creating a unique space to discuss competing theories and viewpoints is core to a GAC. Many scientific questions at heart involve opposing viewpoints, schools of thought, or theories. Often it is just a matter of recognizing them and framing scientific questions in terms of a discourse. The purpose of framing GAC discussions as arbitration of theories or viewpoints is thus to highlight open questions, puzzles, or unresolved challenges rather than to create divides where none exist. Science is full of cases where empirical data don’t quite work with current theories, or where new or modified approaches may be beneficial. Finding these cases and exploring them with the goal of designing novel approaches to resolving such challenges is a core element of the “adversarial” part of a GAC. You can check out previous examples to see how others have framed their debate. 

If you are unsure, whether your idea is suited for a GAC or possibly another format (e.g., the new Community Events proposal format for CCN 2024), feel free to informally reach out to us (gac@ccneuro.org). 

Does a GAC team have to consist of clear proponents of either perspective or theory? 

In many cases, theoretical positions are identified with particular communities and scientists. Sometimes, there are no “pure” proponents of either theory, and often everyone involved will admit that future experiments might prove either side correct.

To expose and highlight the difference between competing theories, members in a GAC may playfully take more accentuated positions than they would usually do. This is of course fully optional, but will make a GAC particularly productive. Importantly, this over-accentuation of positions, if present, can and should be transparent to the audience. Like a designated devil's advocate in a group discussion, for which everyone recognizes their contrarian role, everyone involved knows that GAC members take more polarizing views in service of a productive scientific debate.