Unverified50% confidenceEventExact time

The Best Falsification Award was given to Utkarsh, who reproduced a paper on optimal robustness and learning-augmented paging and discovered errors in both the theoretical and empirical results.

1

Sources

50%

Confidence

Medium-term (~90 days)

Relevance

8/12/2026

First Seen

Valid until: 11/10/2026

Sources

AI智能体复现ICML 2026论文：34%被验证，23%存在问题

Related Claims

UnverifiedThe supervisor of postdoctoral researcher Peter Degen discovered that a paper he published in 2017 had been abnormally cited in large numbers by AI-assisted generated academic papers.57% similar Unverified2017年Pathak等人在ICML会议上发表的好奇心驱动探索论文提出了一种基于预测误差的内在奖励机制54% similar UnverifiedThe paper's author replied and confirmed the existence of the normalization error found by ChatGPT.53% similar UnverifiedA Reddit user who claims to have no mathematics background used ChatGPT to discover a normalization error in two recently published papers on the Riemann Hypothesis.51% similar

Cite This Claim

Stable URI

https://kongchang.com/claim/737195

API

curl https://kongchang.com/api/v1/knowledge/claims/737195

MCP

get_claim(id=737195)