[preprint-dedup-review] 3 unconfirmed preprint-duplicate exclusions need review

The FLoRA preparation pipeline auto-resolved **3** preprint/publication duplicate pair(s), but `confirmed_preprint_duplicates.csv` has not been touched in over 7 days (last edit 43.1 days ago).

Please review the candidates and either confirm the auto-default by adding the rows to `cache/confirmed_preprint_duplicates.csv` with `action=keep_1` / `keep_2`, or override with `keep_both`.

Until each pair is confirmed, the surviving row's `alt_identifier_*` will **not** be annotated with the dropped DOI.

**First 3 of 3 auto-resolved pair(s):**

- `10.1523/eneuro.0150-20.2020` ⇄ `10.31219/osf.io/pzy5q` (replication, sim=1) — auto: auto_keep_1
- `10.1177/0956797620955209` ⇄ `10.31234/osf.io/m6ghr` (replication, sim=0.945) — auto: auto_keep_1
- `10.2466/07.10.17.pr0.107.4.139-144` ⇄ `10.2466/pr0.107.1.139-144` (replication, sim=0.978) — auto: auto_keep_1

Full log: `output/preprint_dedup_candidates.csv`

_Filed automatically by `R/preprint_dedup.R`._