paper-with-me

Papers

When Do Autoregressive Sequence Models Forecast Physical Wavefields? A Controlled Study on Synthetic Seismograms

2026-06-09 · Waleed Esmail, Stuart Russell, Jana Klinge, Alexander Kappes, Christine Thomas arxiv

Long-horizon autoregressive forecasting of oscillatory physical signals, such as seismograms, gravitational-wave strain, and similar wavefields is limited by error accumulation: as a causal model is fed its own outputs over hundreds of steps, small per-step errors compound into phase drift that pointwise metrics fail to detect. We ask when such rollout stays stable, using synthetic three-component seismograms as a physically structured testbed and the \textsc{SeismoGPT} autoregressive forecaster as the model under study. Through controlled architecture ablations evaluated on free rollout with paired significance tests, we isolate the contribution of each design choice. Multi-token prediction is the dominant stabilizer, accounting for almost the entire improvement over a single-token baseline ($+0.042$ median NCC); a horizon-embedding hybrid prediction head and a cross-horizon STFT-magnitude coherence loss each add a small but consistent further gain. Performance depends sharply on a context-ratio threshold near one, roughly the full P-S interval of observed signal, below which rollout generalization collapses. The dominant residual failure is a polarity inversion: a magnitude-based spectral loss reduces its incidence but, by construction, cannot penalize an inverted segment once it occurs, identifying phase-aware objectives as the natural next step. We frame this as a controlled study of rollout stability on oscillatory wavefields, not a benchmark of forecasting architectures.

📄 PDF Abstract BibTeX arXiv:2606.10868

Code (0)

등록된 구현이 없습니다.

Similar Papers 제목 키워드 기반

Data-Driven Forecasting of three-Component Seismograms Using Transformer Architectures

2026-06-01 · Waleed Esmail, Stuart Russell, Jana Klinge, Alexander Kappes 외 arxiv

Forecasting seismic waveforms beyond observed data remains challenging due to the nonlinear, dispersive, and multi-scale nature of seismic wave propagation. In this work, we introduce \textsc{SeismoGPT}, a transformer-ba…

SutraNets: Sub-series Autoregressive Networks for Long-Sequence, Probabilistic Forecasting

2023-12-22 · NeurIPS 2023 11 · Shane Bergsma, Timothy Zeyl, Lei Guo

We propose SutraNets, a novel method for neural probabilistic forecasting of long-sequence time series. SutraNets use an autoregressive generative model to factorize the likelihood of long sequences into products of cond…

Time Series

Poly-Autoregressive Prediction for Modeling Interactions

2025-02-12 · CVPR 2025 1 · Neerja Thakkar, Tara Sadjadpour, Jathushan Rajasegaran, Shiry Ginosar 외

We introduce a simple framework for predicting the behavior of an agent in multi-agent settings. In contrast to autoregressive (AR) tasks, such as language processing, our focus is on scenarios with multiple agents whose…

Autonomous VehiclesPredictionTrajectory Prediction

Searth Transformer: A Transformer Architecture Incorporating Earth's Geospheric Physical Priors for Global Mid-Range Weather Forecasting

2026-01-14 · Tianye Li, Qi Liu, Hao Li, Lei Chen 외 arxiv

Accurate global medium-range weather forecasting is fundamental to Earth system science. Most existing Transformer-based forecasting models adopt vision-centric architectures that neglect the Earth's spherical geometry a…

Weather Forecasting

Explanation of Dynamic Physical Field Predictions using WassersteinGrad: Application to Autoregressive Weather Forecasting

2026-04-24 · Younes Essafouri, Laure Raynaud, Luciano Drozda, Laurent Risser arxiv

As the demand to integrate Artificial Intelligence into high-stakes environments continues to grow, explaining the reasoning behind neural-network predictions has shifted from a theoretical curiosity to a strict operatio…

Weather Forecasting