Murali Annavaram
3 papers in the PaperMetrix corpus
Papers by this author
-
Look Ahead ORAM: Obfuscating Addresses in Recommendation Model Training.
2021 · arXiv (Cornell University)
In the cloud computing era, data privacy is a critical concern. Memory accesses patterns can leak private information. This data leak is particularly challenging for deep learning recommendation models, where data associated with a user …
-
Estimating Privacy Leakage of Augmented Contextual Knowledge in Language Models
2025
Language models (LMs) rely on their parametric knowledge augmented with relevant contextual knowledge for certain tasks, such as question answering.However, the contextual knowledge can contain private information that may be leaked when answering queries, and …
-
DEL: Context-Aware Dynamic Exit Layer for Efficient Self-Speculative Decoding
2025 · arXiv (Cornell University)
Speculative Decoding (SD) is a widely used approach to accelerate the inference of large language models (LLMs) without reducing generation quality. It operates by first using a compact model to draft multiple tokens efficiently, followed …