The silence around ICLR rebuttals this cycle isn't a glitch, it's a signal worth heeding. When a batch of ten to twelve papers sees zero author responses and no reviewer engagement, even on borderline fixable reviews, something systemic is at play. Our take is that the community is experiencing a quiet but real exhaustion with a process that has lost its return on effort, and that has direct consequences for how research gets evaluated and how researchers allocate their limited energy.
Consider what a rebuttal demands. Authors must craft careful, evidence-based responses to critiques that often feel uneven in quality or confidence. As we explored in How Reviewers Confidence Shapes Split Decisions in AI Research, a single weak accept from a low-confidence reviewer can sit alongside a reject from a confident one, leaving authors to guess which voice matters. When the process feels like a lottery where the most articulate rebuttal still loses to a reviewer who simply won't budge, the rational choice for many becomes disengagement. The same dynamics that make split decisions frustrating also make the rebuttal phase feel hollow.
This withdrawal matters practically. If authors and reviewers stop treating rebuttals as a genuine dialogue, the conference loses one of its few mechanisms for correcting errors before acceptance. The Navigating AC-Only Comments at ICLR 2027 post highlights how even the process for private area-chair communication introduces complexity that can discourage participation. When the system adds friction without clear payoff, people vote with their time. The result is a review process that becomes more about surviving noise than about improving science.
The deeper issue is volume. With submissions climbing each year, the pressure on reviewers grows, and the quality of engagement drops. Our piece on Exploring a hard cap on ML conference papers to sharpen research focus outlines a structural fix that could restore meaning to both reviews and rebuttals. A cap would not solve every problem, but it would reduce the noise that makes silence a rational response. The specific takeaway is this: until conferences address the underlying incentive structure, where writing a rebuttal feels like shouting into an empty room, the silence will only spread. Watch for whether ICLR's organizers acknowledge this trend in their post-mortem, because ignoring it means accepting a review process that no one fully trusts.