Anyone submit ML articles to ACM journals (eg. TOPML or TIST)? [D]

A question surfaced in the machine learning community at the start of May that sounds, on the surface, like administrative housekeeping: has anyone actually submitted ML work to ACM journals, specifically TOPML or TIST? The kind of post that typically drowns beneath conference deadlines and arxiv preprint notifications. But the question points at something structural that the field has quietly avoided examining. Machine learning built its prestige architecture almost entirely around conferences — NeurIPS, ICML, ICLR — not journals. A paper that spends eight months in traditional peer review is, in most ML subfields, a paper that arrives late to a conversation that moved on without it. Speed became culture. Culture became identity. And somewhere in that process, ACM's journal pipeline — which in almost every other computing discipline represents the gold standard of scholarly record — became a road that ML researchers rarely take and almost never discuss taking.

The thread's author is asking a practical question with an uncomfortable subtext: are these venues actually functioning for the community they were designed to serve, or have they been quietly bypassed by a field that outgrew the traditional publication timeline? TOPML, ACM's Transactions on Machine Learning, launched as a dedicated home for the discipline. TIST, the Transactions on Intelligent Systems and Technology, has existed longer. Neither is invisible. But neither is on the shortlist that a hiring committee or a grant reviewer typically scans first.

If confirmed as a genuine gap in community knowledge — meaning that experienced ML researchers genuinely have little or no firsthand experience submitting to these venues — the implications reach further than publication etiquette. Journals provide something conferences structurally cannot: extended peer review, revision cycles, and a permanent, version-of-record document. For work that ages well — theoretical foundations, longitudinal studies, methods papers that deserve more than eight pages and a poster — the absence of a healthy journal culture creates an archival problem. Work gets cited from preprints that never formally resolved, from conference papers that compressed a year of insight into a NeurIPS format, from proceedings that are technically peer-reviewed but at a pace and depth that bears little resemblance to what other sciences mean by the term. If ACM's ML journals are functioning but invisible, that's a promotion and community engagement failure. If they're struggling for quality submissions because the field simply doesn't route there, that's a structural wound that compounds over years.

Watch for responses in the thread from researchers who have direct submission experience — acceptance timelines, review quality, editorial responsiveness — because that anecdotal record is currently the only meaningful map of whether these venues are viable options or polite fictions.

NewsHive monitors these sources continuously. All signal titles above link to the original reporting.

Intelligence by NewsHive. Need help navigating what this means for your business? Contact GeekyBee →