- Rating
- 2
- Confidence
- 2
- Informativeness
- 2
- Presentation
- 2
- Soundness
- 2
Read review
Paper summary
GridReS is an online sampler for video post-training that replaces offline deduplication. Clips are bucketed on a PCA grid over embeddings, cells are drawn ∝ n^α, and within a cell the sampler prefers clips far from recent draws. At equal coverage it halves near-duplicate co-exposure R on the Kandinsky-6 pool and lowers it by 25% on OpenVid-100k. A single fine-tuning job gains 1.84 VBench imaging points over random sampling, and per-domain jobs match production deduplication.
Critical strengths
- The paper reports its seed floor and negative results very openly. The analysis in App. B of what region-level samplers can and cannot do is useful on its own.
- Per-draw cost is low, and the code is released.
Critical weaknesses
- The link from R to model quality is not shown. Per domain, random sampling over the full pool also matches production (Table 11), so that result doesn't isolate GridReS. In the single job, GridReS covers only 19.6% of the pool vs. 33.7% for random and is worse on V-JEPA 2 metrics, so the imaging gain may come from which clips were seen. The evaluation also lacks text alignment, audio metrics and human preference.
- E is precomputed and the sampler uses no training signal, so the natural baseline is a simple offline permutation that spreads near-duplicate groups apart. It is missing.
Questions for authors
- How does an offline permutation that spreads near-duplicate groups compare?
- In the single-job run, can you separate the effect of which clips were selected from the effect of their order?
- How do R and coverage behave with more ranks?