2026-06-08原文2 min readUpdated: 2026-06-08

Skip a Layer or Loop It? Learning Program-of-Layers in LLMs

Researchers propose PoLar, a training-free dynamic program that skips or loops pretrained layers in LLMs per input. It achieves better accuracy with fewer layers on math reasoning benchmarks, suggesting fixed-depth inference underutilizes LLM capacity.

SourcearXiv Machine LearningAuthor: Ziyue Li, Yang Li, Tianyi Zhou

[2606.06574] Skip a Layer or Loop It? Learning Program-of-Layers in LLMs

[Submitted on 4 Jun 2026]

Title:Skip a Layer or Loop It? Learning Program-of-Layers in LLMs

View a PDF of the paper titled Skip a Layer or Loop It? Learning Program-of-Layers in LLMs, by Ziyue Li and 2 other authors

View PDF HTML (experimental)

Abstract:Large language models (LLMs) perform inference by following a fixed depth and order, non-recurrent execution of all layers. We reveal the wide existence of training-free, flexible, dynamic program-of-layers (PoLar), where pretrained layers can be packed as modules and then skipped or looped to form a customized program for each input. For most inputs, substantially shorter program executions can achieve the same or better accuracy, while incorrect predictions of the original LLM can be corrected by alternative programs with fewer layers. These observations indicate that inference admits multiple valid latent computations beyond the standard forward pass. To efficiently achieve PoLar in practice, we propose a lightweight PoLar prediction network, which learns to generate execution programs that dynamically skip or repeat pretrained layers for each input. Experiments on mathematical reasoning benchmarks demonstrate that PoLar consistently improves accuracy over standard inference and prior dynamic-depth methods, often while executing fewer layers, and that these gains persist under out-of-distribution evaluation. Our results suggest that fixed-depth execution captures only a narrow subset of an LLM's latent reasoning capacity.

Comments: Accepted at ICML 2026. Substantially extends arXiv:2507.07996. Code: this https URL

Subjects:

Machine Learning (cs.LG)

Cite as: arXiv:2606.06574 [cs.LG]

(or arXiv:2606.06574v1 [cs.LG] for this version)

https://doi.org/10.48550/arXiv.2606.06574

arXiv-issued DOI via DataCite

Submission history

From: Ziyue Li [view email] [v1] Thu, 4 Jun 2026 17:59:58 UTC (574 KB)

Full-text links:

Access Paper:

View a PDF of the paper titled Skip a Layer or Loop It? Learning Program-of-Layers in LLMs, by Ziyue Li and 2 other authors

View PDF

HTML (experimental)

TeX Source

view license

Current browse context:

cs.LG

new | recent | 2026-06

Change to browse by:

References & Citations

NASA ADS

Google Scholar

Semantic Scholar

Data provided by:

Bibliographic Tools

Bibliographic and Citation Tools

Bibliographic Explorer Toggle

Bibliographic Explorer (What is the Explorer?)

Connected Papers Toggle

Connected Papers (What is Connected Papers?)

Litmaps Toggle

Litmaps (What is Litmaps?)

scite.ai Toggle

scite Smart Citations (What are Smart Citations?)

Code, Data, Media

Code, Data and Media Associated with this Article

alphaXiv Toggle

alphaXiv (What is alphaXiv?)

Links to Code Toggle

CatalyzeX Code Finder for Papers (What is CatalyzeX?)

DagsHub Toggle

DagsHub (What is DagsHub?)

GotitPub Toggle

Gotit.pub (What is GotitPub?)

Huggingface Toggle

Hugging Face (What is Huggingface?)

ScienceCast Toggle

ScienceCast (What is ScienceCast?)

Demos

Replicate Toggle

Replicate (What is Replicate?)

Spaces Toggle

Hugging Face Spaces (What is Spaces?)

Spaces Toggle

TXYZ.AI (What is TXYZ.AI?)