NVIDIA: We Need a New Math for AI Models

NVIDIA: The Math Breaking AI Agents. UC Berkeley: Provides first answers. Beijing University: A full immune system for AI. All rights w/ authors: Staying on Task: Testing the Foundations of Long-Horizon Agent Reliability Jeffrey Willette Krishna C. Puvvada Boris Ginsburg NVIDIA WHEN CONTEXT CHANGES: UNDERSTANDING UPDATE FAILURES IN LLMS Junyu Guo Yuchen Fang Shangding Gu junyuguo24@berkeley.edu yc_fang@berkeley.edu shangding.gu@berkeley.edu Costas Spanos James Demmel Javad Lavaei spanos@e