Autoregressive Model

Abstractive
Non-Deterministic Turing Machine with probability

Autoregression is a statistical technique that uses past values to predict future values in a time series. It's a regression of a variable against itself.

The confusing terminology is that Masked Attention is considered opposite to

Bidirectional LM because information about tokens after the current token is not considered. The key point is unidirectionality/bidirectionality

Autoregressive Model Notion

Decoding Strategy

Multilevel Autoregressive Modeling

Continuous Autoregressive Model

Maxime Labonne - Decoding Strategies in Large Language Models

A Guide to Text Generation From Beam Search to Nucleus Sampling

https://mlabonne.github.io/blog/posts/2022-06-07-Decoding_strategies.html

Decoder-only LLM to Encoder-Decoder Transformer

arxiv.org

https://arxiv.org/pdf/2504.06225

Injective fucntion to final hidden state

arxiv.org

https://arxiv.org/pdf/2510.15511

Autoregressive Model

Abstractive Non-Deterministic Turing Machine with probability

Recommendations

Abstractive
Non-Deterministic Turing Machine with probability