# Sequential Neural Probabilistic Amplitude Shaping: Learning the Channel's Language

arXiv: https://arxiv.org/abs/2605.28143 · HTML: https://mlsanity.com/paper/2605.28143/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 13 of its 21 references that are in the map.
- First posted on arXiv 27 May 2026.

## Builds on (most cited first, 13 of 13)

- [Attention Is All You Need](https://mlsanity.com/paper/1706.03762/) (2017)
- [Estimating or Propagating Gradients Through Stochastic Neurons for Conditional Computation](https://mlsanity.com/paper/1308.3432/) (2013)
- [GLU Variants Improve Transformer](https://mlsanity.com/paper/2002.05202/) (2020)
- [Joint Learning of Probabilistic and Geometric Shaping for Coded Modulation Systems](https://mlsanity.com/paper/2004.05062/) (2020)
- [Analysis of Nonlinear Fiber Interactions for Finite-Length Constant-Composition Sequences](https://mlsanity.com/paper/2006.07039/) (2020)
- [New Lower Bounds on the Capacity of Optical Fiber Channels via Optimized Shaping and Detection](https://mlsanity.com/paper/2112.03796/) (2021)
- [Probabilistic Amplitude Shaping and Nonlinearity Tolerance: Analysis and Sequence Selection Method](https://mlsanity.com/paper/2208.03449/) (2022)
- [Sequence-Selection-Based Constellation Shaping for Nonlinear Channels](https://mlsanity.com/paper/2307.13292/) (2023)
- [Perturbation-based Sequence Selection for Probabilistic Amplitude Shaping](https://mlsanity.com/paper/2407.09006/) (2024)
- [Cost-Gain Analysis of Sequence Selection for Nonlinearity Mitigation](https://mlsanity.com/paper/2411.02004/) (2024)
- [Probabilistic Shaping for Nonlinearity Tolerance](https://mlsanity.com/paper/2412.09581/) (2024)
- [Neural Probabilistic Shaping: Joint Distribution Learning for Optical Fiber Communications](https://mlsanity.com/paper/2507.16012/) (2025)
- [Neural Probabilistic Amplitude Shaping for Nonlinear Fiber Channels](https://mlsanity.com/paper/2602.02716/) (2026)

## Abstract (arXiv)

We present the first neural probabilistic amplitude shaping that outperforms existing methods while accounting for all implementation losses, using a block-less, easily implementable sequential autoregressive encoder compatible with arithmetic distribution matching, yielding reduced rate loss and higher achievable information rates.