NeurIPS 2026: Sydney, Atlanta and Paris Conference Guide
Plan for NeurIPS 2026 with verified dates for Sydney and its Atlanta and Paris satellites, research themes, registration choices and a practical paper-reading strategy.
Verified 16 October 2026. This is an independent event guide. Confirmed organizer details are identified below; attendance and preparation suggestions are editorial advice.
Confirmed dates and the three-location format
NeurIPS 2026 is the fortieth Conference on Neural Information Processing Systems. Its main meeting is scheduled for Sydney, Australia, from December 6 to 12. Atlanta, United States, and Paris, France, host satellite meetings from December 9 to 13. The dates are not interchangeable: anyone following an author, workshop or presentation should confirm the assigned location before making travel arrangements. The Sydney venue information identifies the International Convention Centre.
The Sydney timetable separates an expo day on December 6, tutorials on December 7, the main conference on December 8–10, and workshops and competitions on December 11–12. These labels describe different learning opportunities. They should not be treated as a promise that one ticket covers every component.
Research focus and who should attend
The official call covers original machine-learning research across language and multimodal models, vision, robotics, reinforcement learning, optimization, theory and scientific applications. It also includes data-centric and socio-technical work. This makes the event useful to researchers comparing new methods, doctoral students locating related work and engineers examining whether a result can survive the conditions of an actual deployment.
A practitioner should arrive with a research question rather than a shopping list of fashionable model names. For example, a retrieval team might examine whether a paper tests long documents, measures source attribution and reports retrieval errors separately from generation errors. A robotics engineer might ask how simulation results transfer to physical environments. Those are suggested ways to evaluate work, not descriptions of unannounced sessions.
Build a paper-first attendance plan
Start with a small reading map: one familiar problem, one adjacent method and one unfamiliar direction. Read abstracts before allocating time to posters. For each shortlisted paper, record the claimed contribution, baseline, evaluation setting and the strongest reason it might fail outside the reported experiment. A useful poster conversation can then begin with a specific question instead of a request to repeat the abstract.
Keep research maturity separate from publication acceptance. An accepted paper can contribute a valuable insight without supplying a maintained software library or a production-ready system. Look for code availability, computational requirements, licensing and independent reproduction. If a demonstration appears impressive, ask what inputs were excluded and whether the comparison used equal resources. These checks help turn conference enthusiasm into an informed technical decision.
Registration, remote access and practical choices
The organizer's calculator distinguishes student, academic and industrial affiliations and selected meeting components. A virtual-only option is listed for conference livestream access and interaction through the mobile app. Current availability and the complete price should be checked in the official registration flow; this guide does not present an invented flat admission fee. NeurIPS also warns that it has no authorized third-party registration or housing partner outside its official channels.
For an employer-funded visit, write down the evidence you intend to bring back: perhaps an evaluation protocol, three relevant papers and two reproducible experiments. Budget for the whole trip and confirm cancellation conditions before paying. International attendees should verify entry requirements with the relevant government authority; a conference invitation does not establish eligibility to enter a country.
What is still subject to confirmation?
This profile verifies the published dates, broad structure and research remit. It does not guarantee admission, the presence of any individual researcher, or identical programming at the satellite sites. Room assignments, speaker availability and individual sessions can change. Use the official schedule as the operational source, and check the presentation location immediately before committing to a meeting with an author.
NeurIPS 2026 FAQ
Which city is the main conference?
Sydney is the main location. Atlanta and Paris are satellite locations with different date ranges, so select the site that matches the work and people you intend to follow.
Can I attend online?
The pricing page lists a virtual-only pass. Check exactly which streams and interactions it includes before choosing it instead of an in-person registration.
Is this mainly an introductory AI course?
No. Its central purpose is research exchange. A newcomer can prepare by reading background material and choosing tutorials carefully, but should expect technical papers and specialist discussion.
How can I get lasting value from the event?
Reserve time after the meeting to reproduce one result, document an assumption that changed your view and share a short reading list with colleagues. That produces a more useful outcome than an unfiltered collection of slides.
Official sources and image credit
- NeurIPS 2026 official conference — checked 16 October 2026.
- NeurIPS 2026 dates and deadlines — checked 16 October 2026.
- NeurIPS 2026 research scope — checked 16 October 2026.
- NeurIPS 2026 pricing and cancellation policies — checked 16 October 2026.
- NeurIPS 2026 venue information — checked 16 October 2026.
- NeurIPS official media kit — checked 16 October 2026.
Image: NeurIPS Foundation, official conference logo from the NeurIPS media kit. Official source.