Charter

What the network is, who belongs to it, how it is run, and how this document changes.

This is the founding document of the network: what it is, who belongs to it, who coordinates it, how the seminars run, and how the document is amended. It is a draft until the founding labs ratify it; see the note at the end.

1. Name

• Full name: Greater Bay Area Computational and Cognitive Neuroscience Network.
• Short name: GBA-CCN Network. Acronym: GBA-CCN.
• Chinese name, for posters and institutional paperwork only: 粤港澳大湾区计算与认知神经科学网络.

English is the default working language, so that researchers across institutions can take part in the same events. Names may be listed in Chinese and English. Teaching material may be contributed in either language when clearly labelled.

2. Purpose

• Pool the audience that no single department has, so that any member lab can invite a speaker on behalf of a network of labs across the participating institutions rather than on behalf of one department.
• Run a joint seminar series with rotating hosts.
• Keep a public, curated roster of participating labs.
• Index tutorials and teaching material, and list open positions.
• Host summer schools and other multi-day events when ready.

3. What the network is and is not

The network is an agreement among labs. It is not a legal entity, a registered society or a 学会. It has no membership fee, no president and no board, and owns nothing except the files in its public repository. Registering a society in Hong Kong or an association on the mainland carries legal duties the group does not want yet.

Formalisation clause. If the network ever needs a legal form, the decision is taken openly: proposed as a pull request against this charter, open for comment for at least 30 days, agreed by consensus of the coordinators and a majority of member labs, and recorded in the amendment log. Nobody may register an entity in the network’s name without this.

4. Membership

The unit of membership is the lab: a group led by a principal investigator at a university, institute or hospital. Individuals do not join. Students and postdocs take part as audience through the calendar feed and the announcement list; there is nothing to apply for.

• Member lab: a PI-led group at an institution in a Greater Bay Area city: Hong Kong, Shenzhen, Guangzhou, Macau, Zhuhai, Foshan, Dongguan, Huizhou, Zhongshan, Jiangmen or Zhaoqing.
• Affiliate lab: a PI-led group elsewhere, for example in Beijing, Shanghai or abroad, that takes part regularly. Affiliates are listed with a badge.

“Computational and cognitive neuroscience” is read broadly: computational modelling of brain and behaviour; systems and cognitive neuroscience using computational methods; neuroAI and brain-inspired computing; and neural data science.

Membership is light but not passive. A member lab is expected to take part in network activities and to take its turn in the seminar rotation, because the network has no staff and no budget: what it does, member labs do. There is no minimum attendance requirement, and a quiet year is not by itself grounds for removal, though the annual confirmation of an entry still has to be answered; a lab leaves the roster only as section 8 describes, or under the code of conduct.

No lab is expected to share unpublished data, research ideas, funding or other confidential material with the network or with another member lab. What the network asks for is time and attention. The member lab guidelines set out what taking part looks like in practice.

Labs apply through the “[Lab]” issue form, linked from the join page. Intake is automatic; approval is human. A coordinator reviews every application, and nothing appears on the site until the pull request is merged, because the roster is only a credential if it is curated. The entry lists the PI, lab name, institution and city, keywords, links and, with permission, a photo that is displayed on the network’s channels only.

5. Coordinators

The network is run by three to seven coordinators, one per participating institution where possible.

• Term: two years, staggered so that roughly half change each year; consecutive terms are allowed.
• Selection: nominated by member labs, confirmed by consensus of the sitting coordinators; if consensus fails, by a simple majority of member labs.
• Responsibilities: review submissions; keep the seminar calendar and host rotation; hold the speaker nomination queue; maintain the site and repository; answer the contact address on the About page.
• Decisions: by consensus, with a simple majority of coordinators as fallback, recorded in the repository.
• Repository ownership: at least two coordinators hold owner rights at all times, so no single person can lock the network out of its own files.

Until the founding labs confirm the first coordinators, one interim coordinator, Xiangbin Teng (滕相斌) of The Chinese University of Hong Kong, maintains the site and repository. The current list is on the About page.

6. Seminar series

• Slot: alternate Thursdays, 16:00 to 17:00 Hong Kong Time (UTC+8), on Zoom, with Tencent Meeting (VooV) as fallback.
• Rotating hosts: each seminar is hosted by one member lab in turn. The host invites the speaker, chairs, and circulates the meeting link to the announcement list; links are never published on the site.
• Swaps and deferrals: a lab may exchange its slot with another lab, or ask the coordinators to move it to a later round. A lab that swaps tells the coordinators the new order, because they keep the calendar and the rotation. A slot is never left empty without telling them.
• Junior slot: the host may open with a 15-minute talk by a student or postdoc.
• Speaker nominations: anyone may nominate through the “[Speaker]” issue form. The queue is read by the coordinators and never published on the site. Nominations are GitHub issues in a public repository, so they should say only what the nominee could read.
• Recording: only with the speaker’s written consent, given before the talk; recordings go to both Bilibili and YouTube.
• Honorarium: none by default. A host institution may fund travel for an in-person visit.

7. Other activities

• Learning index: the resources page is an index, not a file host. Tutorials stay where their authors host them. Larger material produced by the network, such as summer school recordings, goes in a separate materials repository using GitHub Releases, connected to Zenodo so each edition receives a DOI.
• Positions: member and affiliate labs may list open positions; listings are removed when their closing date passes.
• Events: everything with a date is an event, whether seminar, workshop, summer school, hackathon or journal club. The code of conduct applies to every one of them; coordinators will review it with member labs and add what an in-person event needs before the network’s first multi-day event.

8. Delisting and annual refresh

A lab is removed when its PI asks or leaves the region; in the second case affiliate status is offered. Each year the coordinators ask every listed lab to confirm its entry. Entries unconfirmed after two reminders are removed at the next refresh.

9. Data, consent and openness

Everything the network publishes is a file in a public repository, and every change is a pull request anyone can read. Data is hosted on GitHub, outside mainland China. Labs publish their entries voluntarily, through a consent box on the form. Text content is CC BY 4.0; code is MIT; photographs are displayed with permission and are not licensed for reuse. What is collected, how photos are processed and how a lab withdraws are described in Privacy and consent, which reads Hong Kong PDPO and mainland PIPL in plain terms. Neither that document nor this one is legal advice.

10. Amendments

Anyone may propose an amendment as a pull request against this file. It stays open for comment for at least 14 days, announced on the list, and is merged on the coordinators’ consensus, with a simple majority as fallback. Typo and link fixes skip the comment period.

Draft for ratification by the founding labs. Adopted: not yet.

Amendment log