Lune Research Technical Report
Grounding language models in full-text scientific literature at inference time.
11 min read
Essays and guides on AI-native scientific research.
Grounding language models in full-text scientific literature at inference time.
11 min read
Language models write references from memory, so some of them never existed. What the studies measured, why retrieval alone falls short, and what works.
8 min read
What CORE's A* and CCF's A mean, who decides them, where the two lists disagree on the venues you publish in, and how to use a rank without misjudging a paper.
6 min read
A worked example in Claude Code: sweep top venues, follow citations, read full text, tabulate the methods, then check every sentence against its source.
7 min read
