Evan Almunawar · Zenodo (CERN European Organization for Nuclear Research) 2026 · 2026
DOI: 10.5281/zenodo.23121314
Counts differ because each database indexes a different set of publications. We treat OpenAlex as the canonical count; Google Scholar is not shown (no API, and crawling it violates its ToS).
Siesta is a small-scale independent research project exploring identity coherence in an Indonesian-language orchestrator model — a "brain" that stays behaviorally consistent while delegating specialized tasks to separate hidden models, following an architectural pattern we call Human-Parity Orchestration (HuPO). Built entirely from scratch (not fine-tuned) on limited consumer hardware (GPU/TPU with restricted quotas), this preprint documents the architecture, training challenges, and honestly-reported limitations of an ongoing solo research effort under MEA Ecosystem. Siesta adalah proyek riset independen berskala kecil yang mengeksplorasi konsistensi identitas (identity coherence) pada model orchestrator berbahasa Indonesia — sebuah "otak" yang tetap konsisten secara perilaku sambil mendelegasikan tugas spesialis ke model-model tersembunyi terpisah, mengikuti pola arsitektur yang kami sebut Human-Parity Orchestration (HuPO). Dibangun sepenuhnya dari nol (bukan fine-tuning) dengan hardware konsumen terbatas (GPU/TPU berkuota terbatas), naskah riset ini mendokumentasikan arsitektur, tantangan pelatihan, dan keterbatasan yang dilaporkan secara jujur dari upaya riset solo yang masih berjalan di bawah MEA Ecosystem.
No comments yet — start the discussion below.