Eran Tromer

Computational integrity with a public random string from quasi-linear PCPs.

2016

A party executing a computation on behalf of others may benefit from misreporting its output. Cryptographic protocols that detect this can facilitate decentralized systems with stringent computational integrity requirements. For the computation’s result to be …

You Autocomplete Me: Poisoning Vulnerabilities in Neural Code Completion

2020 · arXiv (Cornell University)

Code autocompletion is an integral feature of modern code editors and IDEs. The latest generation of autocompleters uses neural language models, trained on public open-source code repositories, to suggest likely (not just statically feasible) completions …