Blog
Notes on agents, and the mathematics behind them
Learning notes and research write-ups, organised by topic and series so they can be read in order.
LLM Agents
Multi-agent systems, memory and consultation: whom an agent should trust, and when.
1 post
Mathematical Foundations
Bayesian inference, linear algebra and analysis, worked through from first principles.
Coming soon
Sequence Models
State-space models, Mamba and attention: how sequence architectures work and why.
Coming soon
Paper Reading
Close readings of papers I found worth the time.
Coming soon
LLM Agents
Multi-agent systems, memory and consultation: whom an agent should trust, and when.
Understanding BaRe-Mem
1/3The ideas behind our Bayesian reliability memory, one piece at a time.
- Mathematical Intuition Behind BaRe-Mem
- Agent teams: ranking workers before they answer· planned
- Reading the experiments: robustness to misleading advisors· planned
2026
-
Mathematical Intuition Behind BaRe-Mem
The mathematics behind our Bayesian reliability memory.