Metraipolis › archive › dc_47df3df4ce8e

Follow the Features: SAE Feature-Guided Data-Efficient Instruction-Tuning of Large Language Models — arXiv:2611.04085

Source
arXiv (en)
Original
https://arxiv.org/abs/2611.04085
License (as archived)
CC0-1.0
Archived
2026-11-04T21:51:28.825Z
SHA-256 of the archived text
4fffff24ccbfb8cfbc615b205a567436654c24e745c312e7018c4ba4026ca5c2
Stable URL of this page
https://metraipolis.com/documento/dc_47df3df4ce8e

How to cite

Follow the Features: SAE Feature-Guided Data-Efficient Instruction-Tuning of Large Language Models — arXiv:2611.04085 — arXiv (en), https://arxiv.org/abs/2611.04085 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_47df3df4ce8e (Metraipolis archive dc_47df3df4ce8e, sha256 4fffff24ccbfb8cfbc615b205a567436654c24e745c312e7018c4ba4026ca5c2)
BibTeX
@misc{dc47df3df4ce8e,
  title = {Follow the Features: SAE Feature-Guided Data-Efficient Instruction-Tuning of Large Language Models — arXiv:2611.04085},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2611.04085},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_47df3df4ce8e; sha256 4fffff24ccbfb8cfbc615b205a567436654c24e745c312e7018c4ba4026ca5c2},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2611.04085. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Selecting small, effective instruction-tuning datasets requires identifying which aspects of the original data to retain. We introduce Follow the Features (FtF), a data-selection method that uses sparse autoencoder (SAE) features to characterize training data and guide subset construction. FtF ranks features using dataset-level activation statistics, scores samples by their expression of highly ranked features, and incorporates feature resampling to encourage diversity. The LLM-SAE pair used for selection is independent of the model being fine-tuned, enabling reuse across target models. We evaluate FtF on two model-dataset settings at a training budget of 100 samples with a fixed sequence length of 256 tokens. FtF achieves performance competitive with GRAPE and S2L, with relative rankings varying across settings. On Qwen 3 8B trained on Alpaca, FtF exceeds full-dataset training by 1.59 percentage points in the harmonic mean of an AlpacaEval-based win rate and Open LLM Leaderboard performance, although full-dataset training retains a higher Leaderboard score. These results support SAE feature statistics as a useful signal for data-efficient instruction tuning.

Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.LG, nuovi lavori del giorno).
Categorie arXiv: Machine Learning (cs.LG).
Pagina del lavoro: https://arxiv.org/abs/2611.04085 — 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 2026-11-04T21:38:49.698Z dalla pagina https://arxiv.org/list/cs.LG/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).