agentleFS
Sign inStart free
thistleknotthistleknot/skills/bpe-bm25/SKILL.md

bpe-bm25

Skill0 starsChanged 6 months agoNo licenceView on GitHub

DEFAULT FOR TERM SALIENCE (operator 2026-10-04): a term's salience is its share of the sparsevec INNER PRODUCT between a chunk and its sparsevec neighbours, over a BPE-tokenised vocabulary hosted in Postgres -- not a BM25 score; see "Term salience" below. Also: shrink the lexical vocabulary behind a pgvector sparsevec BM25 index so HNSW can build a connected graph at scale, and measure the cost on recall@50 against gold derived from the corpus itself -- no hand-picked constants, no LLM judge. Use when a sparse index has millions of distinct terms and HNSW has been forced to a low m / ef_construction, and you want to reduce the candidate terms (BPE over a BM25-selected vocabulary) with the vocabulary size chosen by measurement. Trigger on "sparsevec vocabulary too large", "HNSW m=4", "reduce candidate terms", "BPE for lexical search", "raw BM25 vs sparsevec vs BPE", or any plan to pick a vocab size by hand. Companion to `stat-partitioning` (median/MAD bands) and `reduce_overlaps` (chunk sizing).

The file

We split this skill into the rules it gives an agent

5 rules. About 0 are common, 4 are distinctive, and 1 is very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈4

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 1

    Very specific

    Tied to their project's names, files or tools. Probably skip.

No licence file, so all rights are reserved — read it at the source. We count its rules but don't quote them. Read them on GitHub.

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

Check your setup, free