Weksa
b97ec1d048
Prompt::lower prints the item's speaker as a fixed token with its forms ...
The speaker's id never prints; its forms, shared by every item, print under
the constant entity:speaker. A speaker with no form in the language names no
entity and does not refuse. A referent citing the real id entity:speaker is
refused (SpeakerTokenTaken). Docs state what the promise guards.
Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
2026-11-13 17:05:56 +00:00
Research: fine-tuned lowerers and embedding-mapped lexicons
2026-10-29 09:45:11 +01:00
Prompt::lower prints the item's speaker as a fixed token with its forms
2026-11-13 17:05:56 +00:00
park-node-body fixes: remove Discord-derived ontology content and its catalog entries; contract is planned design only; drop stale fixture refs (squash of weksa-lowering/park-node-body-fix 3327edd..HEAD; Soul s1, s2 findings)
2026-10-29 10:41:36 +01:00
Draft Weksa interlingua standard
2026-05-19 14:57:37 +01:00
calibrate: per-arm candidate counts and mild-arm offsets ( #82 )
2026-11-13 14:53:42 +00:00
language-ontology-seed fix s2: neutral ja-JP:1 (no politeness per line), entry text defects, drafter residue out of the seed and score vectors, front-door and schema/tex prose
2026-11-12 15:39:48 +00:00
Entity forms carry a ruled pronunciation; lower --ipa emits it
2026-11-07 12:04:41 +00:00
Document utterance embedding handoff
2026-05-19 17:47:31 +01:00
ja-item19: encoder prompt states one role per predication and the group shape for coordination ( #56 )
2026-11-12 18:48:40 +00:00
Draft Weksa interlingua standard
2026-05-19 14:57:37 +01:00
language-ontology-seed fix s4: ja-JP:6 bans ne-softening and hedges only; front door and lowering docs stop naming a target-language profile as register owner
2026-11-12 16:09:36 +00:00
docs: prior art survey and plan ( #80 )
2026-11-13 12:17:12 +00:00
Draft Weksa interlingua standard
2026-05-19 14:57:37 +01:00
park-node-body: park the Node body and Discord-sourced harness at parked/weksa-node-body; true README (squash of weksa-lowering/park-node-body 2583fb9..e41b60d)
2026-10-29 10:28:33 +01:00
Draft Weksa interlingua standard
2026-05-19 14:57:37 +01:00
scene-in-group: rubric r1 and 98 labelled scenes ( #77 )
2026-11-13 08:24:45 +00:00
Add MiMo speech provider target contract
2026-06-11 19:17:02 +01:00
park-node-body fixes: remove Discord-derived ontology content and its catalog entries; contract is planned design only; drop stale fixture refs (squash of weksa-lowering/park-node-body-fix 3327edd..HEAD; Soul s1, s2 findings)
2026-10-29 10:41:36 +01:00
b97ec1d048