GRaM + CogSci dual submission: continuation patterns in BCI as geometric invariant

[bmorphism]

Two papers, one conservation law

bmorphism/worlds#9 targets GRaM @ ICLR 2026 (Apr 15 deadline). The draft covers six continuations + GF(3) conservation. But the BCI application — the reason this matters beyond PL theory — lives here in horse.

CogSci 2026 (Rio, Jul 22-25) accepts computational modeling papers. A companion submission showing GF(3) conservation in real EEG/fNIRS data would make the GRaM theory paper AND the CogSci empirical paper cite each other.

The dual submission

	GRaM (ICLR)	CogSci
Focus	Six continuation patterns, geometric invariant	Conservation in neural signals, BCI pipeline
Data	Zig tests (6/6 pass)	EEG/fNIRS from Cognionics + OpenNIRScap
Theory	Trit conservation, Heyting algebra	Sheaf descent, cofunctor control
Deadline	Apr 15 AoE	~Feb 2026 (past) OR late-breaking poster

Against other horse issues

• fix: import lc-macros in top-level macros.tree to fix Forester build #1-cleanup: remove yarn.lock, gitignore lock files + build artifacts #4 all build infrastructure — this COMMUNICATES it
• worlds#9 has the theory — horse has the application
• worlds#11 (Gödel dispatch) is the punchline — self-modifying BCI

Implementation

• Extend worlds GRaM draft with BCI section (§8: Application to Neural Signals)
• CogSci abstract from horse-0003 + horse-0004 stage descriptions
• ~500 words added to existing draft

Theory without application is philosophy. Application without theory is engineering. Both together is science.

cc bmorphism/worlds#9 bmorphism/worlds#11 bmorphism/worlds#6