Metraipolis › archive › dc_04b992bb0db4

Query-based Submodular Data Selection for Efficient Instruction Tuning — arXiv:2611.04059

Source
arXiv (en)
Original
https://arxiv.org/abs/2611.04059
License (as archived)
CC0-1.0
Archived
2026-11-04T12:27:38.639Z
SHA-256 of the archived text
e29bfeb5ac5f0f6f79d2841287a2b68bf4f5dd8bef142e1e5f14f9d6b43d8883
Stable URL of this page
https://metraipolis.com/documento/dc_04b992bb0db4

How to cite

Query-based Submodular Data Selection for Efficient Instruction Tuning — arXiv:2611.04059 — arXiv (en), https://arxiv.org/abs/2611.04059 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_04b992bb0db4 (Metraipolis archive dc_04b992bb0db4, sha256 e29bfeb5ac5f0f6f79d2841287a2b68bf4f5dd8bef142e1e5f14f9d6b43d8883)
BibTeX
@misc{dc04b992bb0db4,
  title = {Query-based Submodular Data Selection for Efficient Instruction Tuning — arXiv:2611.04059},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2611.04059},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_04b992bb0db4; sha256 e29bfeb5ac5f0f6f79d2841287a2b68bf4f5dd8bef142e1e5f14f9d6b43d8883},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2611.04059. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

As large language models are fine-tuned for increasingly specialized tasks, their performance depends more on which training examples are selected than on how many. Given a large heterogeneous training pool and a small target set specifying the desired capability, targeted data selection aims to find a compact subset that maximizes downstream performance. Existing methods score each candidate independently against the target and select the top-ranked examples, yet this point-wise strategy ignores inter-example interactions: at tight budgets the selected subset is dominated by near-redundant points that waste capacity on overlapping information. We propose Query-based Submodular functions with Gradient-guided Synthesis (QSGS), which formulates targeted selection as a submodular graph cut over a gradient similarity graph. A modular relevance term pulls selected examples toward the target task, while a pairwise dispersion penalty suppresses intra-subset redundancy. The resulting objective balances target relevance and intra-subset redundancy using the same formulation across different selection budgets. Empirically, the benefit of diversity-aware selection becomes increasingly pronounced as the data budget decreases. To further close coverage gaps that no selection from existing data can bridge, we introduce a gradient-guided synthesis step that identifies under-covered target directions via residual subspace analysis and generates synthetic examples to fill them. Experiments on MMLU, BBH, and TydiQA across Llama-2, Llama-3, and Mistral, show that QSGS at a 0.5% budget matches or exceeds full-data performance on Mistral-7B, outperforms the strongest baseline across all benchmarks, and yields increasing gains as the budget shrinks.

Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.AI, nuovi lavori del giorno).
Categorie arXiv: Artificial Intelligence (cs.AI).
Pagina del lavoro: https://arxiv.org/abs/2611.04059 — 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 2026-11-04T12:26:47.121Z dalla pagina https://arxiv.org/list/cs.AI/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).