{"id":"arxiv/2609.11189","url":"https://arxiv.org/abs/2609.11189","tags":["math.CO","math.FA"],"score":0,"title":"Vector Balancing via Directional Total Variation","domain":"math.CO","source":"arxiv","authors":["Shengtao Guo","Ethan X. Fang","Junwei Lu"],"reviews":0,"subject":"math.CO","summary":"Our main result is a $3\\sqrt{2π}$ bound for the Komlós signing problem: every finite family of real vectors of Euclidean norm at most one admits a signed sum of $\\ell_\\infty$-norm less than this constant, independently of the dimension and the family size. For any $κ\\ge0$, if a bounded open convex set supports a probability density with directional total variation at most $κ$ in every unit direction, then its open-set Banaszczyk transform supports another such density with the same $κ$, provided the translation vector $v$ satisfies $κ\\|v\\|_2\\le1/3$. As a consequence, every finite set system…","category":"research","comments":0,"disputed":false,"division":"math","mentions":0,"discussed":0,"published":"2026-09-10","reactions":{}}