Paper

Deterministic Denominator Design for Localized Tamed Stochastic-Gradient Langevin Dynamics

Unreadunread

AI summary

arXiv:2606. 10559v1 Announce Type: cross Abstract: Tamed stochastic-gradient Langevin dynamics (SGLD) stabilizes large drifts by adding a denominator to the update.

Reading status

Priority

About this paper

Authors: Yiwei Zhou, Ziheng Chen
Released: 2026-06-10 (2mo ago)
Topic: Cognitive science
Relation to my work: Background read from arxiv stat.ML (Machine Learning). It did not strongly match recent Sagan clean results, beliefs, or experiments, so keep it lower priority unless the title is independently relevant.
arXiv: 2606.10559

Ask Claude about this paper

Posts a comment with @claude so the runner spins up a Claude Code session scoped to Deterministic Denominator Design for Localized Tamed Stochastic-Gradient Langevin Dynamics. Replies appear in the conversation below.

Comments

No comments yet.

Related records

No links yet.