Metraipolis › archive › dc_d36dec044c55

DriveFoundry: Co-Evolving Data and Model for Efficient Closed-Loop Iteration in Autonomous Driving — arXiv:2611.11532

Source: arXiv (en)
Original: https://arxiv.org/abs/2611.11532
License (as archived): CC0-1.0
Archived: 2026-11-10T18:41:23.076Z
SHA-256 of the archived text: 2a6ffdaa4d4e8101bf063845d3e2db82c44721b81206c0762178176871341527
Stable URL of this page: https://metraipolis.com/documento/dc_d36dec044c55

How to cite

DriveFoundry: Co-Evolving Data and Model for Efficient Closed-Loop Iteration in Autonomous Driving — arXiv:2611.11532 — arXiv (en), https://arxiv.org/abs/2611.11532 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_d36dec044c55 (Metraipolis archive dc_d36dec044c55, sha256 2a6ffdaa4d4e8101bf063845d3e2db82c44721b81206c0762178176871341527)

BibTeX

@misc{dcd36dec044c55,
  title = {DriveFoundry: Co-Evolving Data and Model for Efficient Closed-Loop Iteration in Autonomous Driving — arXiv:2611.11532},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2611.11532},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_d36dec044c55; sha256 2a6ffdaa4d4e8101bf063845d3e2db82c44721b81206c0762178176871341527},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2611.11532. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

A driving foundation model must learn reusable decision knowledge from rare critical events and apply it under strict runtime constraints. We introduce DriveFoundry, a closed-loop framework improving iteration efficiency under data and computation budgets by coupling structured causal data refinement with adaptive inference. The Data Engine builds traceable scene-candidate records linking scene evidence and factor-decision relations to candidate trajectories and outcome assessments. Candidate comparisons and counterfactual annotations clarify how scene factors affect behavioral feasibility and preference. Verified model failures guide VLM-assisted annotation and expert correction, refining training data and annotation systems. The model engine, Xiaomi-Drive-o1, unifies Perception, Causation, Action, World, and Value within a shared vision-language backbone to enable cross-task knowledge transfer. Block-wise latent refill passes candidate-specific evidence and predicted visual consequences between task blocks, connecting scene understanding, planning, and evaluation. A learned value-of-computation head selects tasks and allocates latent tokens based on expected benefit and cost, adapting inference to each decision. On PhysicalAI, the 9B model achieves the lowest ADE@1 across evaluated splits and horizons among tested models. In AlpaSim, it outperforms Alpamayo 1.5 across all metrics and achieves lower close-encounter and off-road rates than the 34B Alpamayo 2 Super, while trailing in progress and AlpaSim scores. For the 0.8B variant, joint training reduces 5 s minADE@6 by about 54% without altering the inference path or token budget. Adaptive inference matches full reasoning performance using only 28% of tokens. Real-vehicle tests confirm runtime feasibility, supporting a unified paradigm for learning from critical experience and dynamically allocating computation.

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/2611.11532 — 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 2026-11-10T18:37:37.513Z dalla pagina https://arxiv.org/list/cs.AI/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).