4 papers
Tool Use Enables Undetectable Steganography in Multi-Agent LLM Systems
Jimmy Laurence Rippin, Simon C. Marshall, David Demitri Africa +1
Increasingly autonomous agentic AI systems pose novel multi-agent risks, such as secret collusion via covert communication channels. The natural defence to these collusion attempts…
Quantum machine learning advantages beyond hardness of evaluation
Riccardo Molteni, Simon C. Marshall, Vedran Dunjko
The most general examples of quantum learning advantages involve data labeled by cryptographic or intrinsically quantum functions, where classical learners are limited by the infea…
Debate is efficient with your time
Jonah Brown-Cohen, Geoffrey Irving, Simon C. Marshall +3
AI safety via debate uses two competing models to help a human judge verify complex computational tasks. Previous work has established what problems debate can solve in principle,…
Uniform bounds and uncertainty for asymptotics of representations of -adic
Rahul Dalal, Mathilde Gerbelli-Gauthier, Simon Marshall
We prove two results on the growth of dimensions of fixed vectors of representations of -adic under principal congruence subgroups: First, a uniform bound on t…