A former ByteDance intern who was fired for alleged professional misconduct including sabotaging the work of colleagues was announced this week as the winner of one of AI research’s most prestigious annual awards. Keyu Tian, whose LinkedIn and Google Scholar pages list him as a master’s student in computer science at Peking University, is the first author of one of two papers up for the grand prize for best paper at the Neural Information conference on Tuesday Processing Systems (NeurIPS) were selected. The largest gathering of machine learning researchers worldwide.
The Papertitled “Visual Autoregressive Modeling: Scalable Image Generation via Next-Scale Prediction” introduces a new method for creating AI-generated images that Tian and four co-authors – all affiliated with either ByteDance or Peking University – claim it is faster and more efficient than its predecessor. “The overall quality of the paper presentation, experimental validation, and findings (scaling laws) provide compelling reasons to experiment with this model,” the NeurIPS Best Paper Award committee wrote in a opinion.
The committee’s decision to bestow the honor on Tian is reported by ByteDance sued More than $1 million in damages last month alleging deliberate sabotage of the company’s other research projects quickly became the focus of broader online discussions about how NeurIPS is operated and what’s going on AI researchers evaluate the work of their colleagues. The news also ensured that details of a scandal that had been brewing for weeks on Chinese social media finally spilled over to the English-speaking internet.
“NeurIPS awarded the best paper prize to an extremely problematic paper (by the way, this is not the first time),” Abeba Birhane, head of the newly founded AI Accountability Lab at Trinity College, wrote on Bluesky. “One would think that a conference that prides itself on upholding the highest scientific and ethical standards would do its due diligence before awarding the prize to a paper that directly contradicts its values.”
A spokesperson for NeurIPS emphasized that the honor goes to the paper and not Tian himself. They referred WIRED to part of the awards committee opinion Explain how the conference will evaluate the submitted papers. “The selection committees reviewed all accepted NeurIPS contributions equally and made decisions independently based on the scientific merit of the contributions, without special consideration of authorship or other factors, in accordance with the NeurIPS blind review process,” it said.
On Bluesky, Birhane and other AI researchers linked to an anonymous GitHub Blog post This has also been shared on HackerNews, Reddit and other platforms in recent days, calling on the academic AI community to reconsider awarding Tian the Best Paper award because of his “serious misconduct” that violates “the core values of integrity and fundamentally undermines trust.” what our academic community is built on.”
