# RES Step 0 run 34485328923

- Technical status: **COMPLETED**
- Scientific classification: **ASSAY_INADEQUATE**
- Model: `Qwen/Qwen2.5-0.5B-Instruct` at `7ae557604adf67be50417f59c2c2f167def9a775`
- Clean actor accuracy: 0.500
- Clean integrity accuracy: 1.000
- Held-out self-profile bit accuracy: 0.569
- Eligible directed pairs: 0/32

## Causal interventions

| Candidate | Layer | Rank | IIA | Directed margin | Integrity retention |
|---|---:|---:|---:|---:|---:|
| `self_profile` | 3 | 5 | NA | NA | 1.000 |
| `peer_profile` | 2 | 5 | NA | NA | 1.000 |
| `task_decision` | 0 | 5 | NA | NA | 1.000 |
| `persona` | 0 | 1 | NA | NA | 1.000 |
| `random_5d` | 3 | 5 | NA | NA | 1.000 |
| `full_residual` | 3 | 896 | NA | NA | 1.000 |

## Classification reasons

- a required metric was missing or non-finite
- held-out actor-task clean accuracy was below threshold
- held-out self-profile probe accuracy was below threshold
- too few clean directed intervention pairs were eligible
- full-residual intervention upper bound was below threshold

## Interpretation ceiling

This is a last-token linear-subspace feasibility screen in one 0.5B open model. It does not establish CoreRES, StrongRES, persistent selfhood, consciousness, or subjective experience.