@misc{indiciaebb5ef21cf7a7,
  title = {One LR Doesn't Fit All: Heavy-Tail Guided Layerwise Learning Rates for LLMs},
  author = {Di He and Songjun Tu and Keyu Wang and Lu Yin and Shiwei Liu},
  year = {2026},
  url = {https://arxiv.org/abs/2605.22297},
  note = {Source identifier: 2605.22297}
}