Neural Collapse predicts that balanced one-hot classification pushes model representations to be equally far from each other; a symmetric configuration that depends only on the output label and ignores any semantic similarity in the inputs. This creates a puzzle: next-token prediction language models are trained predominantly (as context length increases) with one-hot labels: the same context is very unlikely to appear twice in training with different labels.
arXiv:2606. 15521v1 Announce Type: cross Abstract: Tokenization introduces representational redundancy: under a fixed token vocabulary, every byte string admits many valid token encodings, or segmentations, that decode to the same surface string.
By Kanishk Jain, Matthew Day, Tankut Can
arXiv:2507. 01414v2 Announce Type: replace Abstract: We introduce a new family of toy problems that combine features of linear-regression-style continuous in-context learning (ICL) with discrete associative recall.
By Sultan Daniels, Dylan Davis, Dhruv Gautam, Wentinn Liao, Gireeja Ranade, Anant Sahai
arXiv:2606. 07559v2 Announce Type: replace-cross Abstract: Fine-tuning a language model often fails silently when its correct completion must outrank a near-synonym competitor.
By Vaibhav Prakash, Jayasri Dontabhaktuni
arXiv:2512. 07355v2 Announce Type: replace Abstract: Two traditions of interpretability have evolved side by side but seldom spoken to each other: Concept Bottleneck Models (CBMs), which prescribe what a concept should be, and Sparse Autoencoders (SAEs), which discover what concepts emerge.
By Alexandre Rocchi, Thomas Fel, Gianni Franchi
arXiv:2608. 16269v1 Announce Type: cross Abstract: Recent advances in neural topic models with pre-trained language models (PLMs) have achieved strong performance by leveraging general-domain pre-training, yet their topic interpretability often degrades on specialized corpora.
By Seung-Won Seo, Won Ik Cho, Yongmin Yoo