Home » 2026
Yearly Archives: 2026
Pichay: Treating LLM Context as Virtual Memory
I recently submitted an early version of the work I’ve been doing in exploring the current way AI agents use context window. I learned a number of useful things along the way: First, here’s a link to the preprint I posted recently: [2603.09023] The Missing Memory Hierarchy: Demand Paging for LLM Context Windows The work […]
Premature Collapse is the Root of All Evil
Premature optimization is the root of all evil. – Donald Knuth One of the common, recurring themes that I have observed for more than a year is the tendency to push for simple solutions. I suspect that much of this is because we prefer to have simple unambiguous answers. They require less cognitive effort and […]
What a long strange trip its been
I have realized that I have not been using this forum effectively to talk about my work. My recent posts are all about AI: Epistemic Honesty Revisited: Worse Than I FearedEpistemic Honesty: An Unusual Commodity for Large Language ModelsThe Risks of Using Gemini Code So, I decided to catch up here. PhD Research I spent […]
Epistemic Honesty Revisited: Worse Than I Feared
In my earlier post about Epistemic Honesty I provided an initial snapshot into the behavior of current AI models when asked to summarize a non-existent paper. Subsequent analysis showed that the 62% fabrication figure was overly conservative. This arose because we initially counted a disclaimer as a refusal. This is overly restrictive. When I use […]
Recent Comments