Apple ML

Apple tests lexical substitutions for multilingual transfer Apple Machine Learning Research has described LINK, a data-level method intended to improve multilingual language models when target-language training data is scarce. The approach changes part of the high-resource English pretraining...

Apple tests categorical flows as a non-autoregressive LM path Apple Machine Learning Research says it has scaled Categorical Flow Maps to a 1.7B-parameter language model trained on 2.1T tokens. The work positions continuous diffusion and flow matching as a possible alternative to autoregressive...

Apple researchers test a technical lock for fine-tuning LLM weights Apple Machine Learning Research has published work on DLR-Lock, a method intended to make open-weight language models harder to adapt after release. The paper frames the issue as a tradeoff: shared weights support adoption...