diff --git a/src/content/blog/2026-09-05-frontier-closed.adoc b/src/content/blog/2026-09-05-frontier-closed.adoc index 504fb44..c2e7e5f 100644 --- a/src/content/blog/2026-09-05-frontier-closed.adoc +++ b/src/content/blog/2026-09-05-frontier-closed.adoc @@ -6,12 +6,12 @@ v1.0, 2026-09-05 == Introduction -Two weeks ago the phonological layer shipped with an honest gap: the -Arabic client student scored 8.26 on the full benchmark — a disclosed -miss. This week that gap closed the only way gaps should close: every -rung measured on the full set, confidence intervals on every -separation, and the two causal hypotheses that remained both tested to -a verdict. One verdict was negative. That is the point. +Two weeks ago the phonological layer shipped with a disclosed gap: +the Arabic client model scored 8.26 on the full benchmark. This week +every rung was measured on the full test set, with confidence +intervals on each separation, and the two remaining hypotheses were +both tested. One test came back negative, and that result is published +here with the rest. == The frontier, bracketed @@ -76,8 +76,7 @@ Every leaderboard number we publish can now be re-derived by anyone: The tool reproduces our published verdicts exactly — it re-derives this week's 4.8231 run from its raw predictions and the public benchmark, -intervals included. Protocol-matched comparison should be a command, -not a promise. +intervals included. The comparison is a command, not a claim. == Where this leaves the stack diff --git a/src/pages/about.astro b/src/pages/about.astro index 4faf659..aa41a7d 100644 --- a/src/pages/about.astro +++ b/src/pages/about.astro @@ -137,8 +137,8 @@ const partners: Partner[] = [ run in a web browser.
- Every score, the method used to produce it, and the full results table are on - the neural layer page. + Every score, the method used to produce it, and the full results + table are on the neural layer page.
diff --git a/src/pages/api.astro b/src/pages/api.astro index 6c0795f..6fb4249 100644 --- a/src/pages/api.astro +++ b/src/pages/api.astro @@ -24,8 +24,8 @@ const playgroundSystems = [ The REST API runs athttps://api.interscript.org/v1. No account, no API key.
Send a POST request with a JSON body. Endpoints: /transliterate (one
conversion), /transliterate/batch (many), /systems (the 289-system
- catalogue), /detect (which system explains a source–target pair), and
- /infer (neural models — diacritization and grapheme-to-phoneme). Responses
+ catalogue), /detect (which system explains a source–target pair),
+ and /infer (neural models — diacritization and grapheme-to-phoneme). Responses
allow cross-origin requests, so browser scripts can call the API directly.
Try live transliteration →
Browse {totalSystems} systems
+ Use it in your code
- Without installing anything: the same models answer the REST endpoint
-
- The whole stack runs in your browser: 289 authority systems, plus the
- neural layer for the conversions no committee ever published. Nothing
- leaves the page.
+ Runs entirely in your browser: {289} authority systems and the neural
+ models for text that has no published system. No text is sent to a
+ server.
POST https://api.interscript.org/v1/infer (open CORS, no key), and run in the
+ Without installing anything: the same models answer the REST
+ endpoint POST https://api.interscript.org/v1/infer (open CORS, no key), and run in the
browser at the demo page. In a script with no build step, the map
layer loads from esm.sh — full recipe in the
package's docs/CDN.md.
diff --git a/src/pages/playground.astro b/src/pages/playground.astro
index 0323a1b..20c1377 100644
--- a/src/pages/playground.astro
+++ b/src/pages/playground.astro
@@ -15,9 +15,9 @@ const systems = Object.entries(catalogue as RecordPlayground