What is an average publication outcome for an ML PhD? [D]
A viral Reddit debate reveals typical paper counts for top ML PhDs.
Deep Dive
A Reddit user asks the community: for an ML PhD, is 3–5 first-author papers at top venues like NeurIPS, ICML, ICLR, CVPR, ACL, or EMNLP considered average or above average? The user acknowledges paper count is a crude metric but wants a rough sense of typical outcomes.
Key Points
- The user asked about average first-author papers at venues like NeurIPS, ICML, ICLR, CVPR, ACL, and EMNLP.
- Typical responses suggest 1–3 top-tier first-author papers is a strong outcome, not 3–5.
- Subfield, advisor, and co-authorship norms make publication count a crude metric for PhD success.
Why It Matters
Helps ML PhD students and advisors set realistic expectations for publication productivity.