Adrian Saxton
adriansaxton
AI & ML interests
AI assurance and evaluation
AI security (prompt injection, model misuse, agent permissions)
Agents, tool use and MCP
Human oversight and AI governance
LLM evaluation on real workloads
Retrieval and knowledge graphs for multi-agent systems
Recent Activity
upvoted an article 3 days ago
TorchNeuron: running HF models on AWS Trainium is easier than everOrganizations
None yet