Metraipolis › archive › dc_d54ee08ebacc

From Reweighting to Rewriting: Unlocking the Intervention Effects of Influential Samples in Training Data Attribution — arXiv:2609.02771

Source: arXiv (en)
Original: https://arxiv.org/abs/2609.02771
License (as archived): CC0-1.0
Archived: 2026-10-17T03:38:37.383Z
SHA-256 of the archived text: a4784bd1c62d6f09515ad975b327850fb00b66ddcb21a1ae48d7ad9a27f81be2
Stable URL of this page: https://metraipolis.com/documento/dc_d54ee08ebacc

How to cite

From Reweighting to Rewriting: Unlocking the Intervention Effects of Influential Samples in Training Data Attribution — arXiv:2609.02771 — arXiv (en), https://arxiv.org/abs/2609.02771 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_d54ee08ebacc (Metraipolis archive dc_d54ee08ebacc, sha256 a4784bd1c62d6f09515ad975b327850fb00b66ddcb21a1ae48d7ad9a27f81be2)

BibTeX

@misc{dcd54ee08ebacc,
  title = {From Reweighting to Rewriting: Unlocking the Intervention Effects of Influential Samples in Training Data Attribution — arXiv:2609.02771},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.02771},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_d54ee08ebacc; sha256 a4784bd1c62d6f09515ad975b327850fb00b66ddcb21a1ae48d7ad9a27f81be2},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.02771. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Training data attribution (TDA) aims to identify training examples that shape model behavior, but its intervention value depends on both which examples are selected and how they are modified. Influence functions (IF) estimate behavioral changes under infinitesimal reweighting, yet IF-selected examples often show limited advantages over random selection under conventional weight-based interventions. This raises the question of whether influential examples lack intervention value or whether reweighting fails to realize their behavioral this http URL introduce influence-guided response rewriting, which uses IF to identify intervention targets and replaces their responses with behavior-aligned or behavior-opposed supervision while keeping instructions fixed. Across four open-weight LLMs, we compare rewriting and reweighting on the same influence-selected examples using epistemic abstention as our primary testbed. Response rewriting produces stronger, more persistent, and bidirectional behavioral shifts, while reweighting the same examples yields weak and inconsistent effects. Further analyses show that influence-selected examples provide greater rewriting leverage than alternative selectors, with changes remaining concentrated on target-relevant behaviors. The same qualitative contrast extends to safety refusal. These results distinguish the local reweighting effects captured by influence estimates from the broader intervention leverage of the examples they identify, motivating intervention-aware evaluation of TDA methods.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-03 in cs.AI).
Categorie arXiv: Computation and Language (cs.CL) ; Artificial Intelligence (cs.AI); Machine Learning (cs.LG).
Pagina del lavoro: https://arxiv.org/abs/2609.02771 — 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 2026-10-17T03:13:12.571Z dalla pagina https://arxiv.org/catchup/cs.AI/2026-09-03?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).