Metraipolis › archive › dc_2f253201a2d5

When Less Compute Is More: Adaptive Early Exit Improves Pretrained Outlier Detection — arXiv:2609.32898

Source
arXiv (en)
Original
https://arxiv.org/abs/2609.32898
License (as archived)
CC0-1.0
Archived
2026-10-07T21:00:45.481Z
SHA-256 of the archived text
0a2fe2f834406333adc4cc917a78cdf91bec3b1c11591694ea70add8645635ea
Stable URL of this page
https://metraipolis.com/documento/dc_2f253201a2d5

How to cite

When Less Compute Is More: Adaptive Early Exit Improves Pretrained Outlier Detection — arXiv:2609.32898 — arXiv (en), https://arxiv.org/abs/2609.32898 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_2f253201a2d5 (Metraipolis archive dc_2f253201a2d5, sha256 0a2fe2f834406333adc4cc917a78cdf91bec3b1c11591694ea70add8645635ea)
BibTeX
@misc{dc2f253201a2d5,
  title = {When Less Compute Is More: Adaptive Early Exit Improves Pretrained Outlier Detection — arXiv:2609.32898},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.32898},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_2f253201a2d5; sha256 0a2fe2f834406333adc4cc917a78cdf91bec3b1c11591694ea70add8645635ea},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.32898. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Pretrained tabular foundation models process every dataset at a fixed depth, with inference costs growing with dataset size. To address this, we present the first study of depth-adaptive early-exit for pretrained outlier detection models. While early-exit is typically motivated by efficiency, we uncover a surprising benefit: exiting at the optimal intermediate layer can also improve detection performance on diverse real-world benchmarks by 4.7-7.3% on average, consistent across three distinct foundation models. First, we investigate the factors driving these gains, and identify a key mechanism: context pollution, i.e., the presence of outliers among in-context samples. Our analysis reveals that nearby in-context samples exert increasing influence on query predictions at greater depths, consistent with a retrieval-based view of these models. In effect, early-exit alleviates the adverse effects of retrieving accurate-yet-polluted neighbors, with gains of 13-21% when context pollution matches the natural outlier rate. Motivated by these findings, we pretrain a plug-in router to select a dataset-specific exit layer, using query outlier labels as privileged information available only during router training. The router operates post hoc, leaving the base model parameters and prediction head unchanged. Experiments on three large real-world benchmarks show that, on clean context, the router recovers up to 45% of the oracle gain with up to 1.8x speedup across three pretrained backbones, with larger gains as context pollution increases.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-29 in cs.AI).
Categorie arXiv: Machine Learning (cs.LG) ; Artificial Intelligence (cs.AI).
Pagina del lavoro: https://arxiv.org/abs/2609.32898 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito).
Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina.
Letto il 2026-10-07T20:29:02.495Z dalla pagina https://arxiv.org/catchup/cs.AI/2026-09-29?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).