How to Follow a Top CS Conference Without Attending

You don’t need a conference ticket to benefit from it — a step-by-step system for extracting tutorials, papers, and research directions from any top-tier venue.

Expanded 2025 edition — originally published May 2019

Following conferences strategically — even without attending — is one of the highest-leverage habits a researcher can build. Here’s a systematic approach to extracting maximum learning from top-tier venues.

You don’t need to attend a conference to benefit from it. In fact, most of the value — the accepted papers, tutorials, workshop slides, and keynote talks — is freely available online within weeks of each conference. Learning to mine this information efficiently is a skill that accelerates your research at zero cost.

In this article I’ll walk you through exactly how I follow conferences in my domain, from understanding the Call for Papers to extracting research directions from accepted papers.

Why Following Conferences Matters
Step 1: Reading the Call for Papers
Step 2: Tutorials and Workshop Slides
Step 3: Keynote Talks
Step 4: Proceedings and Accepted Papers
Step 5: Working Notes from Attendees
Building Your Conference Calendar
Tools and Resources

Why Following Conferences Matters

When I was starting my research, I made the mistake of reading papers in isolation — picking up whatever Google Scholar surfaced. The problem: without a sense of what the community cares about, I had no way to judge whether my own ideas were novel or stale.

Following conferences fixes this. It gives you:

1. A current map of research directions. The accepted papers at ACL or NeurIPS represent what the best researchers in the world consider important right now. Reading even the titles and abstracts of a conference proceedings gives you a meaningful update on the field.

2. Role models for your own work. A well-structured ACL paper is a template. Good conferences are full of papers with excellent motivation sections, clean baselines, and honest limitations. Study the format, not just the content.

3. Benchmarks to position your work against. “Our method improves on the ACL 2024 best result by X%” is how papers get accepted. You need to know what the best results are.

4. Future research directions. Most conference papers end with a “Future Work” section. These are effectively a list of research opportunities the authors didn’t have time to explore.

Step 1: Reading the Call for Papers

The CFP (Call for Papers) is one of the most underrated resources for a new researcher. Beyond the submission deadline, it tells you:

Topic areas the conference covers — is your work within scope?
Research themes the community is excited about — often framed as “this year we especially welcome papers on…”
Submission format — long papers (8–10 pages), short papers (4–6 pages), system papers, findings papers
Anonymization policy — most NLP conferences use double-blind review; this affects how you should present your work

Practical tip: Go to the CFP of the 3–4 conferences most relevant to your domain. Print or save the topic list. Before you start writing a paper, verify your work fits at least 2–3 of those topic areas.

Step 2: Tutorials and Workshop Slides

This is the highest-value learning resource at most conferences, and almost all of it is free.

Tutorials are typically 3-hour deep dives into emerging topics, taught by domain experts. They’re designed to get a new researcher up to speed quickly. Examples from recent conferences:

ACL 2024: “Large Language Models for Clinical NLP” (exactly the overlap this site covers)
NeurIPS 2023: “Foundation Models for Medical Imaging”
EMNLP 2023: “Retrieval-Augmented Generation”

Where to find them:

Conference website under “Program” → “Tutorials”
SlidesLive (https://slideslive.com/) — hosts many NeurIPS, ICML, ICLR talks
YouTube channels of major conferences (ACL, NeurIPS, EMNLP all have official channels)

How I use them: I download tutorial slides as PDFs and treat them like textbook chapters for a new topic. An ACL tutorial on clinical NLP is a 60-page summary of 10 years of research, curated by experts.

Step 3: Keynote Talks

Keynote speakers are usually asked to speak about where the field is going, not just where it’s been. This makes keynotes particularly useful for identifying big research questions that aren’t yet well-addressed in the literature.

Good sources:

videolectures.net — large archive of academic talks
SlidesLive — hosts talks from major ML conferences
Conference YouTube channels
Individual researcher pages (many post their keynote slides)

Step 4: Proceedings and Accepted Papers

Reading all accepted papers isn’t realistic, but reading the right 30–40 papers is valuable.

My approach:

Download the full proceedings — ACL Anthology provides PDFs for all NLP papers. DBLP provides links for most CS venues.
Scan all titles — takes 30 minutes. Mark anything that sounds adjacent to your work.
Read abstracts of marked papers — takes 2–3 hours. Reduce to a shortlist of 20–30 papers.
Read introduction + conclusion of shortlisted papers — takes 3–4 hours. This gives you the core contribution and result of each paper.
Deep-read 5–10 papers most directly relevant to your work. These are the papers you need to understand thoroughly and cite.

Tools:

ACL Anthology (https://aclanthology.org/) — complete NLP proceedings
Semantic Scholar (https://semanticscholar.org/) — good for searching with citations
Connected Papers (https://www.connectedpapers.com/) — visual map of related papers

Step 5: Working Notes from Attendees

Some conference attendees publish detailed notes about what they saw. These “working notes” are gold for researchers who couldn’t attend:

They highlight which talks were most impactful
They capture informal discussions that never make it into papers
They often include photos of poster sessions

Where to find them:

Twitter/X during and after conferences — search #ACL2025 or #NeurIPS2024
Medium and personal blogs — search “[Conference name] 2025 notes”
LinkedIn posts from researchers at the conference

Building Your Conference Calendar

For a researcher in AI/ML and medical AI, I recommend following:

Must follow (domain-defining):

ACL, EMNLP, NAACL — if doing NLP
NeurIPS, ICML, ICLR — if doing general ML
MICCAI — if doing medical imaging
AMIA Annual Symposium — if doing clinical NLP and health informatics

Should follow:

AAAI, IJCAI — general AI
KDD — data mining and applied ML
ML4H Workshop (NeurIPS) — medical AI specifically

Track via:

Subscribe to the conference mailing list / newsletter
Follow the conference on Twitter
Set a Google Alert for “[Conference name] accepted papers”

Tools and Resources

Tool	Use
WikiCFP (wikicfp.com)	Deadlines and CFPs
ACL Anthology	NLP proceedings
Semantic Scholar	Paper search + citation graph
Connected Papers	Visual paper relationship map
SlidesLive	Conference talk recordings
Papers With Code	Papers + code + leaderboards
Elicit (elicit.org)	AI-assisted literature review

Personal Motivation

Attending and presenting your work at a conference is one of the major highlights of the journey to becoming an independent researcher. Be it a national or an international venue, conferences allow you to present your work before a large audience.

The journey from a research project idea to paper acceptance and finally oral or poster presentation at a conference is long and arduous, accompanied by a steep learning curve.

So, the first step can be to follow top-tier conferences by going through the list of accepted papers, keynote talks, and tutorials. It has many advantages:

It tells you the relevant topics of interest for the research community and future research directions; you can choose the topic you like
Most of the resources are freely available and are thus accessible to you at no extra cost
Great role models to copy from, for example, presentation skills, how to reply to questions from the audience regarding your paper or presentation

Case Study

1. Workshops and tutorials slides

I found these presentations an easy-to-understand resource for understanding the overall breadth of a field you are new to or an emerging field. I will try to provide a list of such resources, which I found mostly from Twitter :

Tips and tricks for effective Ph.D. by Dr. Renata Borovica-Gajic
Frontiers of Natural Language Processing by Sebastian Ruder, Herman Kamper, Panellists, Leaders in NLP, Everyone

2. Video lectures of oral presentations

videolectures.net and Youtube are two primary resources

SlidesLive has a decent library of interesting conference talks (including some of the ICLR talks)

3. Conference Proceedings

Use DBLP and Google Scholar to get the proceedings of the conferences. It provides an overview of the primary themes and topics of discussion.

4. Working notes

After the conference is over, look out for student notes or some other attendee to the conference within a few days. A good habit is to follow the graduate students and faculty of reputed institutions on Twitter. Choosing people from the same domain as your research interest helps a lot.

ICLR 2019 notes by David Abel

If you want to add anything interesting, please comment below. I would love to hear from you.

Discover more from Medical AI Insights

Subscribe to get the latest posts sent to your email.

How to Follow a Top-Tier Computer Science Conference: A Researcher’s Guide