Metraipolis › archive › dc_56ebe2a20fc3
MaRK: Markov-adapted Recurrent Kernels for Dynamic Operator Conditioning in State Space Models — arXiv:2610.09092
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2610.09092
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-08T11:10:33.922Z
- SHA-256 of the archived text
e7faf004e784b34841b3e03bb50d33b9dfbe31a2b22f7d709abce467974d2840- Stable URL of this page
https://metraipolis.com/documento/dc_56ebe2a20fc3
How to cite
MaRK: Markov-adapted Recurrent Kernels for Dynamic Operator Conditioning in State Space Models — arXiv:2610.09092 — arXiv (en), https://arxiv.org/abs/2610.09092 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_56ebe2a20fc3 (Metraipolis archive dc_56ebe2a20fc3, sha256 e7faf004e784b34841b3e03bb50d33b9dfbe31a2b22f7d709abce467974d2840)
BibTeX
@misc{dc56ebe2a20fc3,
title = {MaRK: Markov-adapted Recurrent Kernels for Dynamic Operator Conditioning in State Space Models — arXiv:2610.09092},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2610.09092},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_56ebe2a20fc3; sha256 e7faf004e784b34841b3e03bb50d33b9dfbe31a2b22f7d709abce467974d2840},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.09092. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
State Space Models (SSMs) offer an efficient alternative to Transformers for sequence modeling, yet conditioning pre-trained SSMs for iterative generation typically operates outside the recurrent operator, through input injection or activation modulation. While such mechanisms expose the model to conditioning information, they leave the underlying temporal dynamics fixed. We introduce MaRK (Markov-adapted Recurrent Kernels), a dynamic operator-conditioning framework that maps context vectors directly into bounded modulations of a frozen SSM's recurrence ($A$), read-in ($B$), read-out ($C$), skip ($D$), and discretization ($\Delta$) parameters. Viewed through the lens of LPV-SSM systems, MaRK induces a context-indexed family of Markov parameter sequences, allowing each diffusion timestep to reshape the model's input-output memory kernel. We instantiate MaRK on a frozen 111M-parameter Hydra SSM backbone and study three adapter geometries: Hypernet, Chebyshev polynomial, and Discrete Cosine Transform kernels. Since these adapters modify the Markov parameter sequence through low-rank auxiliary maps on the frozen backbone, parameter-efficient fine-tuning arises as a structural consequence of the adaptation mechanism itself, requiring only 6.3--11M trainable auxiliary parameters to transition from a bidirectional objective to an iterative diffusion regime. The bounded recurrence parameterization further yields an analytic Affine Quadratic Stability certificate for the modulated recurrence. Through synthetic LPV recovery experiments and Markov-operator diagnostics, we show that MaRK recovers coordinate-invariant temporal operators under matched assumptions and produces distinct, stable timestep-conditioned memory profiles. Empirically, the Chebyshev variant yields the strongest performance, achieving an average validation loss of 2.55, followed by the DCT (2.59) and Hypernet (3.77) geometries. Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.AI, nuovi lavori del giorno). Categorie arXiv: Machine Learning (cs.LG) ; Artificial Intelligence (cs.AI). Pagina del lavoro: https://arxiv.org/abs/2610.09092 — 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 2026-10-08T10:14:06.048Z dalla pagina https://arxiv.org/list/cs.AI/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).