IAAIAdaptive Depth Diagnosis
Adaptive Depth in Looped Transformers: Diagnosing Learned Halting Gates and Trajectory Readouts
Separates the two jobs of a learned halting distribution, choosing which state to read and weighting per-depth losses, on synthetic tasks and Ouro checkpoints. Fixed-prior depth supervision yields difficulty-aware trajectories, and simple post-hoc confidence readouts often match learned gates.
Inside the method
Simplified conceptual schematic. Consult the paper for the complete architecture.
- Recurrence family
- Adaptive depth
- Depth control
- Learned or post-hoc halting
- KV / state strategy
- See paper
Reading note
Failures stem mainly from the trajectory induced by joint gate training rather than gate expressivity. Ouro's ponder gates are competitive but not uniformly Pareto-optimal; measured latency confirms that lower average exit depth saves time.
Sources checked 2026-10-28. This catalog does not imply independent reproduction.
Related works
Cite this work
@misc{haltingtrajectoryreadout2026,
title = {Adaptive Depth in Looped Transformers: Diagnosing Learned Halting Gates and Trajectory Readouts},
author = {Andrei Cristian Popescu and Haitz Sáez de Ocáriz Borde and Pietro Liò},
year = {2026},
eprint = {2607.20519},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2607.20519}
}