Dynamic Chunking for End-to-End Hierarchical Sequence Modeling

N/A

Dynamic Chunking for End-to-End Hierarchical Sequence Modeling Sukjun Hwang Brandon Wang Carnegie Mellon University Cartesia AI sukjunh@cs.cmu.edu brandon.wang@cartesia.ai Albert Gu…

saved by

• Krish Mehta

related reading

• H-Nets - the Past | Goomba Labgoombalab.github.io
• transformer_attention.pdfarxiv.org
• H-Nets - the Future | Goomba Labgoombalab.github.io
• [2512.23675] End-to-End Test-Time Training for Long Contextarxiv.org
• H3: Language Modeling with State Space Models and (Almost) No Attention · Hazy Researchhazyresearch.stanford.edu
• [2309.10668] Language Modeling Is Compressionarxiv.org
• [2212.14052] Hungry Hungry Hippos: Towards Language Modeling with State Space Modelsarxiv.org
• 2502.11089arxiv.org
• https://aclanthology.org/P19-1285.pdfaclanthology.org
• Hyena Hierarchy: Towards Larger Convolutional Language Models · Hazy Researchhazyresearch.stanford.edu
• [2005.11401] Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasksarxiv.org
• arxiv.org/pdf/2512.24880#page=3.56arxiv.org