AI Paper Atlas

Research map / Probabilistic and scientific modeling / Variational and generative sampling

Tail Annealing for Heavy-Tailed Flow Matching

Jean Pachebat

ICML 2026 · Probabilistic and scientific modeling › Variational and generative sampling · approach: Flow matching and flow sampling · 0 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
apply soft-log transform before flow matching (Log-FM vs TTFfix)flow matching · Gumbel+Gaussian, tail index 1.5, d=20median W1 (Pareto margins): 0.963 -> 0.451Improved
apply soft-log instead of arcsinh transformLog-FM · Gumbel+Gaussian, tail index 1.5, d=20median W1 (Pareto margins): 0.550 -> 0.451Improved
use Log-FM instead of gTAF at high dimensiongTAF · Gumbel+Gaussian pooled, d=100median W1 (Pareto margins): 0.704 -> 0.233Improved
use Log-FM for Gaussian margins vs TTFTTF · Gumbel+Gaussian pooled, d=50median W1 (Gaussian margins): 0.030 -> 0.080Hurt
add Hill-gated per-coordinate transform (leave light-tailed margins untouched) (ablation)Log-FM uniform · Gumbel d=20 tau=0.5, tail index 1.5median W1 (Gaussian margins): 0.074 -> 0.043Improved
add Hill-gated per-coordinate transform (ablation)Log-FM uniform · Gumbel d=20 tau=0.5, tail index 1.5median W1 (Pareto margins): 0.449 -> 0.468No clear effect
add Hill gating at tail index 2.5 (ablation)Log-FM uniform · Gumbel d=20 tau=0.5, tail index 2.5median W1 (Pareto margins): 0.068 -> 0.084Hurt
add sampling clamp c=5 (ablation)Log-FM · Gumbel d=20 tau=0.5, tail index 1.5relative W1 Pareto change (%): -2.0%No clear effect
increase Euler solver steps from 10 to 500 (ablation)Log-FM · Gumbel d=20 tau=0.5, tail index 1.5W1 (Pareto margins): 0.521 -> 0.588No clear effect
use Log-FM instead of TTFfix on Student-t dataTTFfix · Student-t benchmark, d=10, nu=1.5W1: 1.78 -> 0.63Improved
use Log-FM on lighter-tailed Student-t (nu=5)TTFfix · Student-t benchmark, d=10, nu=5W1: 0.12 -> 0.14Hurt
use Log-FM instead of TTFfix on real finance dataTTFfix · Fama-French 5W1: 0.449 -> 0.133Improved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy