Talks, posters, and the conference itself

Acceptance email, camera-ready done, flights booked. Now the part nobody teaches: the conference. Here is what actually happens and how to get the most out of it.

The talk

If your paper got an oral slot, you will give a talk, usually 10 to 20 minutes plus questions. The math is brutal and freeing: you cannot present the paper. You can present one idea, one key figure, and one reason to go read the paper.

• One takeaway sentence the audience should remember
• Roughly one slide per minute, fewer is fine
• Big fonts, big figures, minimal text
• Rehearse out loud at least three times, once for timing alone
• Prepare for the two most likely questions

Build the deck in LaTeX so your notation matches the paper: Beamer basics gets you there.

The poster session

Most accepted papers present as posters, and poster sessions are secretly the best part of a conference: two hours of direct conversation with exactly the people who chose to stop at your work. Design one legible from two meters away, and prepare three versions of your pitch: the 30-second version for walkers-by, the 2-minute version for the interested, and the deep dive for the person who read your paper on the plane. See conference posters for the LaTeX side.

Lightning talks and recordings

Two formats that surprise first-timers:

• Lightning talks: some venues give every poster a 1-to-2 minute teaser slot. One slide, one claim, “come find me at poster 42.”
• Recordings: many conferences require a pre-recorded video of your talk, uploaded weeks before the event, with its own deadline. Check the presenter instructions email carefully; this deadline hides in paragraph four.

Networking is the actual point

The talks are recorded. The papers are online. The one thing you cannot download later is the hallway: conversations between sessions, questions at your poster, dinner with people who cite the same papers you do. Some of those conversations become collaborations, postdoc offers, and lifelong colleagues.

If networking sounds terrifying, use the built-in cheat code: ask people about their work. Researchers love talking about their work, and “I saw your poster, how did you handle X?” is a complete social strategy. Attend talks slightly outside your area too; that is where new ideas come from.

Artifacts and code release

Expectations have shifted: reviewers and readers increasingly expect the code, data, or artifacts behind a paper to be public. Many venues run artifact evaluation tracks with badges. Practical version:

• Public repository with a README that says how to reproduce the main results
• Pin dependency versions
• Link the repo from the paper and the poster
• License the code explicitly

If your project already lives in Git with GitHub sync (see sync with GitHub), the release is mostly a cleanup pass, not a migration.

After the conference

Post the final version where people can find it, update your website, and share the one-sentence takeaway. Then rest. The next lesson, your first paper, end to end, puts this whole journey on one map.