Predicting Cable Dynamics with Physical Attention Bias

Hugging Face Daily Papers Papers
physics-simulation attention-mechanism deformable-objects robotics deep-learning simulation

Summary

该论文提出一种物理注意力偏置（physical attention bias），在注意力logits上加入基于弧长距离和欧氏距离的可学习项，以改进对未训练过的电缆等可变形线性对象（DLOs）的运动预测，显著减少长期rollout中的漂移与误差。

Learned simulators for deformable linear objects (DLOs) such as cables have to predict the motion of cables they were not trained on and stay stable over long rollouts. Most of their error occurs where the cable touches itself or the floor. Attention over all pairs of cable segments can represent contact between parts of the cable that are far apart along its length, but attention has no notion of geometry. A cable has two pairwise distances, which agree only while it is straight: the arc-length distance along the cable, which governs elastic forces, and the Euclidean distance in space, which governs contact. We add a physical attention bias, an additive term on the attention logits with a learned rate, and ask which distance it should use. We compare no bias, each distance alone, and both distances on disjoint sets of heads, keeping the rest of the model and the training protocol fixed. A physical bias improves prediction on unseen cables. The gain is largest when attention is the only mechanism that connects distant segments: there, the arc-length bias reduces prediction error by 15% and more than halves the drift in segment length. The Euclidean bias alone stays close to unbiased attention, while assigning both distances across heads is best or near-best on every metric we report. Code and per-run records: https://github.com/avihaig/dlogps.
Original Article
View Cached Full Text

Cached at: 10/09/26, 10:23 PM

Paper page - Predicting Cable Dynamics with Physical Attention Bias

Source: https://huggingface.co/papers/2610.11975

Abstract

Learnedsimulatorsfordeformablelinearobjects(DLOs)suchascableshavetopredictthemotionofcablestheywerenottrainedonandstaystableoverlongrollouts.Mostoftheirerroroccurswherethecabletouchesitselforthefloor.Attentionoverallpairsofcablesegmentscanrepresentcontactbetweenpartsofthecablethatarefarapartalongitslength,butattentionhasnonotionofgeometry.Acablehastwopairwisedistances,whichagreeonlywhileitisstraight:thearc-lengthdistancealongthecable,whichgovernselasticforces,andtheEuclideandistanceinspace,whichgovernscontact.Weaddaphysicalattentionbias,anadditivetermontheattentionlogitswithalearnedrate,andaskwhichdistanceitshoulduse.Wecomparenobias,eachdistancealone,andbothdistancesondisjointsetsofheads,keepingtherestofthemodelandthetrainingprotocolfixed.Aphysicalbiasimprovespredictiononunseencables.Thegainislargestwhenattentionistheonlymechanismthatconnectsdistantsegments:there,thearc-lengthbiasreducespredictionerrorby15%andmorethanhalvesthedriftinsegmentlength.TheEuclideanbiasalonestaysclosetounbiasedattention,whileassigningbothdistancesacrossheadsisbestornear-bestoneverymetricwereport.Codeandper-runrecords:https://github.com/avihaig/dlogps.

View arXiv pageView PDFGitHub0Add to collection

Get this paper in your agent:

hf papers read 2610\.11975

Don’t have the latest CLI?curl \-LsSf https://hf\.co/cli/install\.sh \| bash

Models citing this paper1

#### avihaig/dlogps-checkpoints Graph Machine Learning• Updatedabout 2 hours ago

Datasets citing this paper1

#### avihaig/dlogps-cables Updatedabout 3 hours ago

Spaces citing this paper0

No Space linking this paper

Cite arxiv.org/abs/2610.11975 in a Space README.md to link it from this page.

Collections including this paper0

No Collection including this paper

Add this paper to acollectionto link it from this page.

Similar Articles

HLA-WM: Hybrid Linear Attention for Long-Horizon Video World Models

Hugging Face Daily Papers

HLA-WM提出一种训练免费的混合线性注意力框架，通过几何引导的检索与递归线性状态重组合来缓解GatedDeltaNet在长时视频世界模型中的远距离遗忘问题，在60秒基准上以12倍内存节省同时提升所有一致性与相机控制指标。