Table talks are informal discussions across the community, open to trainees, early-career researchers, and senior researchers alike. In 2026, the current OSR program includes 4 table talks focused on practical questions in open science. All times on this page are shown for Bordeaux, using the Europe/Paris timezone. During the June 15-18, 2026 program, that means CEST (UTC+2).
OHBM 2026 program
Table Talk 1: The Buffet of Open Science
| When: 08:00-09:00 CEST (UTC+2) | June 15, 2026 (Monday) |
Open science offers many practices, tools, and standards, but that breadth can be as intimidating as it is empowering. This session frames open science as a buffet: there is no single plate everyone must take, but there are useful starting points, common ingredients, and practical ways to build a workflow that fits a project’s goals, resources, and constraints.
This table talk will:
-
Give attendees a clearer overview of the stages of a neuroscience project where open-science practices apply.
-
Help participants evaluate where they currently stand and which practices are most realistic to adopt next.
-
Highlight common tools and standards that can make science more open, reproducible, and collaborative.
Speakers:
- Rotem Botvinik-Nezer, Assistant Professor — Hebrew University of Jerusalem
- Guillaume Sescousse, Researcher in cognitive neuroscience — Inserm, France
- Oscar Estebean, Associate Professor — HES-SO Valais-Wallis, Switzerland
- Roberto Toro, Researcher in neuroscience — Pasteur Institute, Paris, France
Moderated by:
- Eva Van Heese — Amsterdam University Medical Center
- Gaëlle Leroux — Lyon Neuroscience Research Centre
Table Talk 2: What's/who's missing in the Era of Large Data?
| When: 13:00-14:00 CEST (UTC+2) | June 15, 2026 (Monday) |
Large-scale neuroimaging datasets have transformed the field, but they also raise harder questions about representation, harmonization, bias, and generalizability. This round table asks which populations, environments, and forms of variation are missing from the datasets that increasingly shape contemporary neuroscience, and how current analytic choices may unintentionally reinforce those gaps.
This table talk will:
-
Identify which populations are overrepresented and underrepresented in major large-scale neuroimaging datasets.
-
Examine how harmonization, normalization, and modeling choices can remove or distort meaningful variation.
-
Explore strategies for making future large-data efforts more inclusive, equitable, and scientifically valid.
Moderated by:
- Felix Hoffstaeder — Institute for Neuroscience and Medicine/Brain and Behavior
- James Kent — University of Texas at Austin/Psychology
Table Talk 3: Preregistration Enhances Creativity in Science
| When: 14:30-15:30 CEST (UTC+2) | June 15, 2026 (Monday) |
Preregistration is increasingly encouraged as part of an open and transparent research culture, yet it is still surrounded by misunderstandings about rigidity, paperwork, and lost flexibility. This session focuses on the common perception that preregistration limits creativity and contrasts that with the ways it can actually clarify ideas early, improve study design, and create a transparent structure for both confirmatory and exploratory work.
This table talk will:
-
Identify common misconceptions about preregistration, including concerns about flexibility, workload, and creativity.
-
Discuss how preregistration can coexist with both confirmatory and exploratory research.
-
Share practical ways to use preregistration without limiting creative scientific work.
Speakers:
- Marta Topor, Assistant Professor — Linköping University, Sweden
- Justine Épinat-Duclos, Research engineer — National Centre for Scientific Research (CNRS), France
Moderated by:
- Eva Van Heese — Amsterdam University Medical Center
- Gaëlle Leroux — Lyon Neuroscience Research Centre
Table Talk 4: Using LLMs to Lower the Barrier to Entry to Open Science
| When: 08:00-09:00 CEST (UTC+2) | June 16, 2026 (Tuesday) |
Large language models are arriving at a moment when many research groups are already strained by the complexity of open-science workflows. This session examines where LLMs can genuinely reduce friction, such as drafting documentation, translating code, assisting with preregistrations, or helping users navigate standards, while also addressing the risks of hallucination, opaque provenance, and over-automation.
This table talk will:
-
Identify productive uses of LLMs within research and open-science workflows.
-
Discuss why human-in-the-loop workflows are essential for responsible AI-assisted research.
-
Compare real-world failure modes and guardrails for using LLMs in scientific work.
Speakers:
- Alejandro de la Vega, Research Assistant Professor — UT Austin
Moderated by:
- James Kent — University of Texas at Austin/Psychology
- Johanna Bayer — Donders Institute for Brain, Cognition and Behaviour