الملخص
Autonomous robots on one-shot missions run over horizons far longer than the trajectories seen during training, under a fixed onboard compute budget. We present FRANK, a 507K-parameter recurrent architecture that combines tau-gated recurrent modules, content-addressable memory, and a feedforward reflex pathway. We evaluate it against recurrent, state-space, and reduced modular baselines at matched parameter count on four algorithmic sequence tasks, trained at length 5-20 and evaluated out to two million tokens. At 100,000x the maximum training length, 6 of 10 FRANK seeds retain exactly 100.0% accuracy, while none of the 50 baseline configurations does, five architectures at ten seeds each with none left incomplete (Fisher exact, two-sided p = 4.2E-6. Targeted lesions across the four tasks yield four distinct component-reliance profiles, consistent with task-dependent allocation across the recurrent, memory, and reflex pathways. Separately, a FRANK policy trained in simulation drives a physical ground vehicle to commanded waypoints through obstacles without teleoperation.
الكلمات المفتاحية
الموضوع
بيانات النشر
- المجلة
- غير متاح
- وصول مفتوح
- وصول مفتوح أخضر
اقتبس هذه المقالة
APA 7
Thornton, I., Shey, A., & Su, W. (2026). Extreme Length Generalization in a Compact Recurrent Architecture for One-Shot Exploration. https://omanscience.com/ar/articles/extreme-length-generalization-in-a-compact-recurrent-architecture-for-one-shot-exploration
MLA 9
Thornton, Izen, et al. "Extreme Length Generalization in a Compact Recurrent Architecture for One-Shot Exploration." https://omanscience.com/ar/articles/extreme-length-generalization-in-a-compact-recurrent-architecture-for-one-shot-exploration.
شيكاغو (المؤلف–التاريخ)
Thornton, Izen, Aaron Shey, and William Su. 2026. "Extreme Length Generalization in a Compact Recurrent Architecture for One-Shot Exploration." https://omanscience.com/ar/articles/extreme-length-generalization-in-a-compact-recurrent-architecture-for-one-shot-exploration.
هارفارد
Thornton, I., Shey, A. and Su, W. (2026) 'Extreme Length Generalization in a Compact Recurrent Architecture for One-Shot Exploration', Available at: https://omanscience.com/ar/articles/extreme-length-generalization-in-a-compact-recurrent-architecture-for-one-shot-exploration.
فانكوفر
Thornton I, Shey A, Su W. Extreme Length Generalization in a Compact Recurrent Architecture for One-Shot Exploration. https://omanscience.com/ar/articles/extreme-length-generalization-in-a-compact-recurrent-architecture-for-one-shot-exploration
IEEE
I. Thornton, A. Shey, and W. Su, "Extreme Length Generalization in a Compact Recurrent Architecture for One-Shot Exploration," https://omanscience.com/ar/articles/extreme-length-generalization-in-a-compact-recurrent-architecture-for-one-shot-exploration.