2026
an archive of posts from this year
| Oct 21, 2026 | Matryoshka Language Model Suites in a thread |
|---|---|
| Oct 21, 2026 | Editing the methods of |
| Oct 20, 2026 | critique of matryoshka paper (GA26) |
| Oct 16, 2026 | Week 0 in one page: tokens keep knowledge apart |
| Oct 16, 2026 | Three things to remember, and where the paper hides them |
| Oct 16, 2026 | The paper in four feeds |
| Oct 16, 2026 | The core idea in one picture |
| Oct 16, 2026 | Rewriting the abstract: lead with the surprise |
| Oct 16, 2026 | Reimplementing CLeq: what does 12.6% actually mean? |
| Oct 16, 2026 | Where this paper comes from: a literature map |
| Oct 16, 2026 | Five follow-ups worth doing |
| Oct 16, 2026 | Fictive facts: how to know exactly what a model saw |
| Oct 16, 2026 | Critique: a clean diagnosis, a partial cure |
| Oct 16, 2026 | A better Figure 2: the threshold the bar chart hides |