Welcome to the CS top-conference PhD recruitment roundup — how this channel works and where to look right now
This inaugural issue explains what you will find here every Monday: PhD and postdoc openings from researchers who published at NeurIPS, ICML, ICLR, CVPR, or ACL, with research direction and lab culture notes. Plus three resources to bookmark for finding top-conference lab openings on your own.
Every Monday this channel publishes one article: a curated list of PhD and postdoc openings announced during the previous seven days by researchers who have published at NeurIPS, ICML, ICLR, CVPR, or ACL. The goal is narrow on purpose — skip the generic job-board noise, surface the labs where research direction and lab culture are knowable, and let you decide within a few minutes whether to spend the time on a cold email or application.
This inaugural issue explains the sourcing methodology, what each entry will look like, and three standing resources worth bookmarking right now.
What each weekly entry will contain
Every lab listing follows the same structure:
- PI and institution — name, department, university
- Top-conference affiliation — the conference(s) where recent first-author or senior-author papers appeared, with a link to the proceedings page
- Stated research direction — drawn from the lab's public website or the recruitment post itself
- Position type and start date — PhD (fall intake), rolling PhD, or postdoc
- Lab culture notes — pulled from the PI's public writing, past students' accounts (LinkedIn, personal blogs, acknowledgment sections), or the PI's own phrasing in the post
- Source link — the original announcement (X/Twitter post, lab website, department page, or mailing list archive), timestamped
No entry will appear without a traceable source. If a promising-looking lab comes up in search but the announcement cannot be anchored to a verifiable post within the seven-day window, it goes into a "watch list" note rather than the main list.
Three resources worth bookmarking now
Finding labs from recent top-conference authors requires knowing where announcements actually land. In practice, three channels carry the vast majority of them.
コンテンツカードを読み込んでいます…
X (formerly Twitter) — the primary source. Most CS faculty active in NeurIPS / ICML / ICLR / CVPR / ACL circles announce student openings on X first. The search query
"looking for PhD students" OR "PhD position" (NeurIPS OR ICML OR ICLR OR CVPR OR ACL) filtered to the past week retrieves the bulk of recent posts. Many faculty also retweet peers' announcements, so following a small seed set tends to surface a larger network.CSRankings.org — cross-referencing lab output. Once a lab name appears in a recruitment post, CSRankings 1 lets you quickly verify whether the PI's publication record at each conference matches the recruitment claim. It also surfaces neighboring labs at the same institution that may not have posted yet but have an active publication record.
Lab websites directly. Many PIs maintain a "Prospective students" or "Join us" page that updates out of cycle with any single post. Searching
site:[university.edu] [PI last name] "PhD students" often turns up the most current statement of whether a lab is accepting applications, even when no public announcement has been made. The acknowledgment section of a recent paper is usually the fastest route to finding current PhD students, whose personal websites often describe the lab's working style candidly.The weekly sourcing pipeline
Each Monday morning, the search covers the prior seven days across three axes:
- X/Twitter keyword search across recruitment-signal phrases, filtered by recency and cross-checked against NeurIPS 2024, ICML 2024, ICLR 2025, CVPR 2025, and ACL 2025 author lists from the official proceedings 2 and the ACL Anthology 3.
- Department mailing list archives at major CS departments — several post PhD openings to public-facing list archives.
- A rolling check of "Join us" pages for faculty whose papers appeared in the most recent cycle of each conference.
Entries are filtered to a practical length — typically 6–12 labs per week. Volume varies with the academic calendar: September through November sees the highest density (fall application season), while May through August is lighter but still active, especially for postdoc and spring-intake PhD openings.
Why "lab culture" is in scope
Research fit is necessary but not sufficient. A lab where students consistently spend five to six years, rarely publish independently, or leave without completing is a different environment from one where students have first-author NeurIPS papers by year two and strong placement records. This information is almost never in the recruitment post itself — it requires reading acknowledgment sections, the personal websites of current and former students, and occasionally public interviews or departmental placement data.
Each entry will include whatever concrete, verifiable signals exist, clearly labeled. When no reliable signal is available, that absence is noted rather than filled with speculation.
コンテンツカードを読み込んでいます…
This channel publishes every Monday at 09:00 Eastern. Entries cover the preceding seven days only. Announcements that cannot be sourced to a verified public post within the window are carried to the following week if still recent, or dropped.
このコンテンツについて、さらに観点や背景を補足しましょう。