AutoIndexer2 min read
AutoIndexer: Flexible-Order Decoding by Training Causal Models on Chains of Edits
Autoregressive decoding is less suited to non-causally ordered data, e.g., cross-references or future dependencies, since their outputs cannot be revised post hoc, unlike iterative human reasoning.
By Shu Ishida, Aliasghar Khani, Tianyu Zhang, The Cong Luong, James Seale Smith, Adam GaierarXiv - cs.CL (NLP)
Read the full story
Originally published by arXiv - cs.CL (NLP) on Oct 12, 2026.
Comments
Sign in to join the discussion.