Search
On March 26, Google Research released a paper detailing the TurboQuant algorithm, asserting that it could substantially decrease AI memory consumption. However, Jianyang Gao, a postdoctoral researcher at ETH Zurich, raised allegations of academic misconduct against it. Gao highlighted that TurboQuant failed to address its notable resemblances to the RaBitQ algorithm, which he published in 2024. Furthermore, he argued that TurboQuant misrepresented the theoretical findings of RaBitQ and intentionally set up an biased experimental setting. Despite repeated attempts by Gao's team to communicate these concerns, Google did not make comprehensive amendments. The publication of the Google paper led to a sharp decline in memory chip stocks, motivating Gao to publicly present concrete evidence (go public with concrete evidence) to defend academic integrity.
-
C114 Communication Network -
Communication Home
