Skip to content

docs(board): E-SPRITE-IPB-HELIX-1 — sprite/helix mapping onto the x265 I/P/B grammar #705

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
AdaWorldAPI merged 1 commit into from
Jul 16, 2026
Merged

docs(board): E-SPRITE-IPB-HELIX-1 — sprite/helix mapping onto the x265 I/P/B grammar #705

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
7 changes: 7 additions & 0 deletions .claude/board/EPIPHANIES.md
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,3 +1,10 @@
## 2026-07-16 — E-SPRITE-IPB-HELIX-1 — moving object = HHTL-anchored splat sprite + helix motion code, mapped onto the x265 I/P/B grammar (crawl-first probe spec'd)

**Status:** CONJECTURE [S/H] + PROBE SPEC (operator-directed amortization; the initial test deliberately replays the x265 GOP grammar, NOT H.268 — "not reaching for the stars").
**What:** the next nested-inverse-pyramid 4×4 amortization: **I-frame** = the sprite's Gaussian-splat set at its HHTL anchor (HEEL|HIP|TWIG); **P-frame** = ONE helix direction code per sprite (`ResidueEdge` 24-bit hemisphere / `Signed360` 48-bit signed — object-level motion replacing x265's per-block MV field, its dominant encoder cost); **B-frame** = parametric interpolation along the helical path between anchors (bidirectional weights, no stored motion). Motion search → address arithmetic; the sprite footprint re-rasterizes through the Morton cascade at certificate-selected LOD; the same tile substrate serves codec rasterization and shader dispatch (E-H268-REPLAYABLE-TILE-1). Second amortization: the minimal wgpu replay harness this probe needs IS the harness PROBE-GPU-LUT is gated on — build once, both consume. Fisher-z canon applies to the read lanes (helix = the 2z rung).
**Scope guards:** NOT x265 bitstream/byte parity (GOP *grammar* replay on our primitives; x265 itself at most an external bitrate/PSNR reference, never a gate); GPU render-grade only, bit-exactness on the CPU/wasm sprite-state path (C9 discipline). **KILL band:** if object-level helix motion collapses back into a dense per-splat MV field on the test paths, the sprite amortization dies as stated.
**Cross-ref:** ndarray plan `x265-sprite-replay-probe-v1.md` (the spec + pass/KILL bands; §5 probe-queue row added), E-H268-REPLAYABLE-TILE-1, E-FISHERZ-CANONICAL-COSINE-REPLACEMENT-1, helix `ResidueEdge`/`Signed360`, ndarray `splat3d` (EWA + depth_cascade), PROBE-GPU-LUT (shared harness). Execution wave = a next plateau after TD-BGZ-TENSOR-PRE-LANE-REVIEW lands.

## 2026-07-16 — E-FISHERZ-CANONICAL-COSINE-REPLACEMENT-1 — THE canonical cosine/centroid replacement = the analytic Fisher-z codec (FamilyGamma, no materialization); helix : Fisher-2z :: cosine-replacement : Fisher-z

**Status:** OPERATOR RULING (2026-07-16) + FINDING for the grounding identity. No layout changed; the L4 semantics are sharpened, not altered.
Expand Down