3 papers
cs.LG2025
Detecting Post-generation Edits to Watermarked LLM Outputs via Combinatorial Watermarking
Liyan Xie, Muhammad Siddeek, Mohamed Seif +2
Watermarking has become a key technique for proprietary language models, enabling the distinction between AI-generated and human-written text. However, in many real-world scenarios…
cs.LG2025
Towards Understanding Text Hallucination of Diffusion Models via Local Generation Bias
Rui Lu, Runzhe Wang, Kaifeng Lyu +3
Score-based diffusion models have achieved incredible performance in generating realistic images, audio, and video data. While these models produce high-quality samples with impres…
math.NT2025
Local divisor correlations in almost all short intervals
Javier Pliego, Yu-Chen Sun, Mengdi Wang
Let be natural numbers, and let denote the -fold and -fold divisor functions, respectively. We analyse the asymptotic behavior of the sum $\sum_{x<n\l…