Phase 4: Plan native evaluation and process pairing #308

Open
opened 2026-10-20 15:04:40 +02:00 by erikinkinen · 0 comments
Owner

Plan ID: P4-20-B
Record type: Implementation leaf
Scientific protocol: 2.2 · Backlog revision: 2.2.1
Specification: sections 8.5; parent requirements and common section-11/17.4 gates apply.
Source: AES_PHASES_2_6_PLAN.md, issue heading at line 6425.
Plan SHA-256: 8acdaeb52d6a94fcad7de64d64521cd465f608344d63560b45f9202f5af71212
Protocol snapshot: Download the complete governing plan.
Tracking parent: P4-20 (#306).

Goal

Separate cost calibration from the measured representation comparison.

Applicable parent requirements

  • P4-20.R1: Generate the 900 calibration-process and 12960 evaluation-process manifests.
  • P4-20.R2: Implement warmups, paired randomized order, medians and immutable coefficient freezing.

Prerequisites

Steps

  • Add the 12960-process matrix and seeded cell-order randomization.

Definition of Done

Every cell has 30 paired fresh processes with the required warmups/lifecycles.

Execution and acceptance contract

Apply the governing semantic/data/statistical contracts in sections 3–5 and the relevant phase section, plus the section-11 checks applicable to this task. Section 17.4 defines the required acceptance evidence. Store the outcome and immutable prerequisite/input/output hashes at

artifacts/backlog/2.2.1/P4-20-B/acceptance.json.

Only explicit dependency and parent-closure edges govern readiness; numbering adds no predecessor dependency. Checkboxes are work/acceptance items, not commit quotas. Keep implementation changes independently reviewable using the existing PR template. Execution, review and gate tasks close with their evidence/decision records and a PR only when tracked artifacts change.

An unsupported scientific effect is a valid result. Missing correctness validation, invalid inference or absent required review is not completion. Do not change endpoints, matrices, margins, inference rules or sample sizes outside the protocol change-control process.

<!-- aes-plan-id: P4-20-B; backlog-revision: 2.2.1 --> **Plan ID:** `P4-20-B` **Record type:** Implementation leaf **Scientific protocol:** 2.2 · **Backlog revision:** 2.2.1 **Specification:** sections 8.5; parent requirements and common section-11/17.4 gates apply. **Source:** `AES_PHASES_2_6_PLAN.md`, issue heading at line 6425. **Plan SHA-256:** `8acdaeb52d6a94fcad7de64d64521cd465f608344d63560b45f9202f5af71212` **Protocol snapshot:** [Download the complete governing plan](https://git.erikinkinen.fi/attachments/31cb768b-4788-4e38-aba9-9d7f0d1ee16d). **Tracking parent:** [P4-20 (#306)](https://git.erikinkinen.fi/erikinkinen/AES/issues/306). ## Goal Separate cost calibration from the measured representation comparison. ## Applicable parent requirements - **P4-20.R1:** Generate the 900 calibration-process and 12960 evaluation-process manifests. - **P4-20.R2:** Implement warmups, paired randomized order, medians and immutable coefficient freezing. ## Prerequisites - [P3-27 (#262)](https://git.erikinkinen.fi/erikinkinen/AES/issues/262) - [P4-19 (#298)](https://git.erikinkinen.fi/erikinkinen/AES/issues/298) ## Steps - [ ] Add the 12960-process matrix and seeded cell-order randomization. ## Definition of Done Every cell has 30 paired fresh processes with the required warmups/lifecycles. ## Execution and acceptance contract Apply the governing semantic/data/statistical contracts in sections 3–5 and the relevant phase section, plus the section-11 checks applicable to this task. Section 17.4 defines the required acceptance evidence. Store the outcome and immutable prerequisite/input/output hashes at `artifacts/backlog/2.2.1/P4-20-B/acceptance.json`. Only explicit dependency and parent-closure edges govern readiness; numbering adds no predecessor dependency. Checkboxes are work/acceptance items, not commit quotas. Keep implementation changes independently reviewable using the existing PR template. Execution, review and gate tasks close with their evidence/decision records and a PR only when tracked artifacts change. An unsupported scientific effect is a valid result. Missing correctness validation, invalid inference or absent required review is not completion. Do not change endpoints, matrices, margins, inference rules or sample sizes outside the protocol change-control process.
erikinkinen added this to the Phase 4 milestone 2026-10-20 15:04:40 +02:00
Sign in to join this conversation.
No milestone
No project
No assignees
1 participant
Notifications
Due date
The due date is invalid or out of range. Please use the format "yyyy-mm-dd".

No due date set.

Reference
erikinkinen/AES#308
No description provided.