yogsoth-ai Factor Removal Strategy
Ablation study approach: systematically remove each factor and observe conclusion stability.
Method
- factor-enumeration lists all factors supporting the conclusion
- single-factor-removal removes one factor at a time
- counterfactual-scenario-construction reasons about the modified scenario
- fragility-measurement scores how much the conclusion degrades
- load-bearing-identification ranks factors by impact of removal
- Optional: flip-point-detection for partial removal (dose-response)
Budget Table
| Parameter | S | M | L |
|---|---|---|---|
| Factors removed | 5 | 10 | 20 |
| Removal iterations | 1 | 2 | 3 |
| Combination removals | 0 | 3 | 8 |
Orchestration
factor-enumeration → [rank by suspected importance]
→ [for each factor]:
single-factor-removal
→ counterfactual-scenario-construction
→ fragility-measurement
→ [if budget allows, test combinations]:
single-factor-removal (multiple factors)
→ counterfactual-scenario-construction
→ load-bearing-identification (final ranking)
Subagents
- factor-enumeration (factor listing)
- single-factor-removal (ablation)
- counterfactual-scenario-construction (scenario reasoning)
- fragility-measurement (degradation scoring)
- load-bearing-identification (ranking)
- flip-point-detection (dose-response, optional)
Available Tactics
Optional, no fixed order; the final leaf is always a sop.
| Tactic | When to use |
|---|---|
| minimal-change-search | Tactic: Generate candidate changes, detect flip-points where conclusion reverses, measure fragility as distance to nearest flip. |
| systematic-factor-ablation | Tactic: List all factors, remove one at a time, assess conclusion stability, rank factors by load-bearing importance. |
Available SOPs
Optional, no fixed order; the final leaf is always a sop.
| SOP | When to use |
|---|---|
| counterfactual-scenario-construction | Construct precise, internally consistent counterfactual scenarios where specified factors are altered, then reason about the resulting conclusion. |
| factor-enumeration | List all key factors, conditions, and assumptions that support or enable the artifact's conclusion. |
| flip-point-detection | Find the minimal change magnitude along a dimension that causes the conclusion to flip from true to false. |
| fragility-measurement | Compute a fragility index from flip-point distances and degradation scores, summarizing how robust the conclusion is. |
| load-bearing-identification | Identify which factors are "load-bearing walls" — factors whose removal would collapse the conclusion. |
| single-factor-removal | Remove one specified factor from the artifact's support structure and reason about how the conclusion changes. |