Metraipolis › archive › dc_5876e475e02f
Support-Set Target Leakage in Relational Foundation Models during In-Context Learning: Model Dependence and Evaluation Reliability — arXiv:2609.36417
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2609.36417
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-07T16:22:36.289Z
- SHA-256 of the archived text
59e7e95d774789697b31aad4ff23fdbf10c8f51592e11a951a5e37a18ae07cb1- Stable URL of this page
https://metraipolis.com/documento/dc_5876e475e02f
How to cite
Support-Set Target Leakage in Relational Foundation Models during In-Context Learning: Model Dependence and Evaluation Reliability — arXiv:2609.36417 — arXiv (en), https://arxiv.org/abs/2609.36417 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_5876e475e02f (Metraipolis archive dc_5876e475e02f, sha256 59e7e95d774789697b31aad4ff23fdbf10c8f51592e11a951a5e37a18ae07cb1)
BibTeX
@misc{dc5876e475e02f,
title = {Support-Set Target Leakage in Relational Foundation Models during In-Context Learning: Model Dependence and Evaluation Reliability — arXiv:2609.36417},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2609.36417},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_5876e475e02f; sha256 59e7e95d774789697b31aad4ff23fdbf10c8f51592e11a951a5e37a18ae07cb1},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.36417. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Relational in-context learning (ICL) uses labeled support examples and their linked relational context to predict labels for new queries. This creates a failure mode when target-derived features are present in the support context but unavailable for the query. We study this setting as support-set target leakage. We construct 20 controlled target-derived features that vary in signal fidelity, representation, semantic transparency, coverage, and zero-, one-, and two-hop relational placement, and evaluate them across 13 RelBench tasks and five relational ICL configurations that vary the ICL head, message-passing depth, pretraining cohort, or relational encoder architecture. We evaluate matched 0-hop, 1-hop, and 2-hop leakage settings, together with a Full leakage condition containing all 20 leaker columns. Within the tested configurations, target-table (0-hop) and Full leakage produce the largest aggregate deviations from clean evaluation, while higher-hop effects are often weaker, consistent with differences in effective exposure associated with temporal reachability, sampling, and aggregation fidelity. Leakage effects are strongly task- and model-dependent and can reverse relative conclusions between model variants even when aggregate changes are small. For leaker detection, we compare an Integrated Gradients (IG)-based screening method with mutual information (MI) and leave-one-column-out (LOCO) on a common Baseline subset. Ranking quality is strongest in the high-impact 0-hop and Full leakage conditions, but detector-based removal does not consistently restore the clean evaluation. A four-task rel-salt case study further shows the same evaluation concern with native-schema leakage candidates from the original relational schema. These results identify the support/query information boundary as an important component of reliable relational ICL evaluation. Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-30 in cs.AI). Categorie arXiv: Artificial Intelligence (cs.AI). Pagina del lavoro: https://arxiv.org/abs/2609.36417 — 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 2026-10-07T15:54:53.498Z dalla pagina https://arxiv.org/catchup/cs.AI/2026-09-30?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).