There's a quiet but telling signal in the experience shared by the researcher who submitted to ICML, ICLR, and TMLR this year. They noticed something that many in the field suspect but rarely say aloud: the quality of feedback at TMLR consistently outshines the big-name conferences, even when the timeline is comparable. The reviews at TMLR felt informed, precise, and genuinely useful. The reviews at ICML, by contrast, often felt rushed, low-confidence, or needlessly harsh without offering a path forward. That difference is not a minor annoyance. It is the difference between a process that helps you improve your work and one that simply filters it.
What stands out here is not that TMLR is perfect. It is that the bar for useful feedback is clearly achievable, and yet the major conferences are not meeting it. The researcher notes that TMLR reviewers are more aware of the topic, ask reasonable questions, and raise concerns where they are warranted. That sounds like the baseline expectation for peer review, not a luxury. But the contrast with ICML suggests otherwise. When reviewers are stretched thin, incentivized to produce quick judgments, or shielded from accountability, the result is exactly what this person describes: feedback that feels performative rather than constructive. For researchers, this is not just frustrating. It changes how they choose where to submit, how they revise, and whether they trust the process at all.
The practical takeaway is straightforward. If you are deciding where to send your work, the venue's reputation matters less than the quality of the engagement you will actually receive. A review that challenges your assumptions with specific, informed questions is worth more than a rejection from a name-brand conference that offers little more than a few dismissive sentences. The researcher's observation suggests that the field is quietly re-evaluating what counts as a meaningful review experience. That shift is not about nostalgia for smaller venues or anti-conference sentiment. It is about recognizing that feedback is the core value of the entire exercise. If the big conferences cannot deliver that, researchers will continue to notice, and they will vote with their submissions.
What this means in practice is that the pressure is now on ICML, NeurIPS, and ICLR to justify their position. They can keep relying on scale and prestige, but the evidence from this year's cycle shows that the cracks are visible. Reviewers are not inherently worse people. The system simply does not reward the kind of care that TMLR demonstrates. Until that changes, the honest answer to the researcher's question, "Are the big conferences even worth it?" is increasingly, "Only if you value the name over the feedback." That is a trade-off each researcher will have to make. But the fact that this comparison is even worth drawing says something important about where the field is headed.