{"slug": "mechanism-aware-ensemble-conditioning-for-data-limited-emulation-of-extreme", "title": "Mechanism-Aware Ensemble Conditioning for Data-Limited Emulation of Extreme Events", "summary": "A mechanism-aware conditioning framework that injects ensemble-covariance statistics into an otherwise unchanged backbone improves rare-event emulation in data-limited chaotic systems, according to arXiv paper 2609.30746v1. In a low-dimensional benchmark, the ensemble covariance directions co-activated with OTD modes and FiLM conditioning cut 99th-percentile exceedance-frequency errors versus an identical no-context Transformer baseline, while in two-layer quasi-geostrophic flow a fixed ensemble-conditioned FiLM-STORN model trained on only 50 time units outperformed an unconditioned STORN trained on the same data and, on averaged high-threshold exceedance diagnostics, beat a baseline STORN trained with 20 times more high-resolution data. The authors present local instability geometry as an actionable conditioning signal for data-efficient rare-event emulation rather than a post hoc interpretation.", "body_md": "arXiv:2609.30746v1 Announce Type: new \nAbstract: Extreme events in chaotic systems are difficult to learn from short trajectories because they are controlled by transient finite-time instability rather than by frequently observed bulk dynamics. We propose a mechanism-aware conditioning plug-in framework that turns a nudged coarse ensemble into a non-intrusive sensor of local instability geometry. In the small-noise regime, the ensemble covariance aggregates the same finite-time deformation kernels that govern local instability, providing a Jacobian-free proxy for the local amplification structure around a synchronized coarse trajectory. A small FiLM module injects statistics of this ensemble geometry into an otherwise unchanged backbone while leaving the coarse simulator unchanged. We demonstrate this interface in two distinct pipelines: a Transformer-style residual-attention corrector for a controlled low-dimensional chaotic system and a probabilistic recurrent STORN corrector for topographic two-layer quasi-geostrophic (QG) flow. In the low-dimensional benchmark, ensemble covariance directions co-activate with OTD modes and FiLM conditioning improves 99th-percentile exceedance-frequency errors over an identical no-context Transformer baseline. In QG, a fixed ensemble-conditioned FiLM-STORN model trained on only $50$ time units substantially improves long-horizon rare-event statistics in the data-limited regime, including density-tail errors, exceedance frequencies, and spatial exceedance-area distributions relative to an unconditioned STORN trained on the same data; on averaged high-threshold exceedance diagnostics, it also outperforms the baseline STORN trained with $20$ times more high-resolution data. These results show that local instability geometry is not merely interpretable post hoc, but an actionable conditioning signal for data-efficient rare-event emulation.", "url": "https://wpnews.pro/news/mechanism-aware-ensemble-conditioning-for-data-limited-emulation-of-extreme", "canonical_source": "https://www.machinebrief.com/news/mechanism-aware-ensemble-conditioning-for-data-limited-emula-jkw6", "published_at": "2026-09-29 04:00:00+00:00", "updated_at": "2026-09-29 05:19:26.723109+00:00", "lang": "en", "topics": ["machine-learning", "ai-research", "neural-networks"], "entities": ["arXiv", "FiLM", "STORN", "Transformer", "quasi-geostrophic flow", "OTD modes"], "also_reported_by": [], "alternates": {"html": "https://wpnews.pro/news/mechanism-aware-ensemble-conditioning-for-data-limited-emulation-of-extreme", "markdown": "https://wpnews.pro/news/mechanism-aware-ensemble-conditioning-for-data-limited-emulation-of-extreme.md", "text": "https://wpnews.pro/news/mechanism-aware-ensemble-conditioning-for-data-limited-emulation-of-extreme.txt", "jsonld": "https://wpnews.pro/news/mechanism-aware-ensemble-conditioning-for-data-limited-emulation-of-extreme.jsonld"}}