TurtLES: a 14 TB benchmark of 400 turbulent 3D wakes for neural PDE surrogates
SorbonneISIR · hf · 2026-11-14
Sorbonne ISIR released TurtLES, a 14 TB benchmark dataset for turbulent 3D neural PDE surrogates.
- 400 trajectories of 400 time steps each, 3–11M points per frame, from high-fidelity implicit large-eddy simulations
- Covers 4 shape families, Reynolds numbers 1000–3900, angles of attack 0–45°, in the shear-layer transition regime
- Combines physically realistic turbulence with active energy cascade, dense irregular meshes, and long-horizon dynamics
- Benchmarks SOTA neural operators on autoregressive rollouts, identifying failure modes: temporal instability, loss of small-scale energy, difficulty on large irregular domains
A new standard for evaluating surrogates on industrially resolved 3D turbulent wakes.
More from Research
- Vauss claims 5.4x faster Gaussian Splat training than gsplat, scene in 106s — ravisparikh · 2026-11-14
- Gradio launches agent challenge: reproduce NeurIPS 2026 papers on Hugging Face, Nov 20–Dec 5 — Gradio · 2026-11-14
- ICLR 2027 Rebuttal Radar tracks which review scores move during the rebuttal window — 3scorciav · 2026-11-14
- HKUST's ODE evolves RL prompts from agent rollouts, boosting multimodal search accuracy — 机器之心 · 2026-11-14
- Paper argues 'The End of Harness Engineering': loops rewrite agent harnesses, humans just write the exam — blaizedsouza · 2026-11-14
- Paper proposes 1500-word briefs plus machine-verifiable appendices to fix AI-era peer review — WalterReade · 2026-11-14