Exa is a modern AI search engine with SERP API, website crawler tools, and deep research API. Power your app with web search AI and web crawling API.
Publication

NITP: Next Implicit Token Prediction for LLM Pre-training

May 24, 2026 · 6 authors · 4 topics

Standard Next-Token Prediction (NTP) super vises language models solely through discrete labels in the output logit space. We argue that this sparse, one-hot supervision leaves the la tent representation space under-constrained, al lowing hidden states to drift into degenerate and anisotropic configurations that limit generaliza tion. To address this issue, we propose Next Im plicit Token Prediction (NITP), which augments discrete prediction with dense, continuous super vision directly in the representation space. NITP requires the model to predict the implicit seman tic content of the next token, using shallow-layer representations from the same model as stable self-supervised targets. Theoretically, we show that NITP regularizes the optimization landscape by mitigating under-constrained degrees of free dom and enforcing a compact, structured repre sentation geometry. Empirically, across dense and MoE models ranging from 0.5B to 9B pa rameters, NITP consistently improves down stream performance with negligible computa tional overhead. Notably, on the 9B MoE model, NITP achieves a 5.7% absolute improvement on MMLU-Pro, along with gains of 6.4% on C3 and 4.3% on CommonsenseQA, with ∼2% ad ditional training FLOPs and no additional infer ence cost. Our implementation is available at https://github.com/aHapBean/NITP.

Showing the abstract — retrieve the full paper via the Exa API.

Xiangdong ZhangDebing ZhangShaofeng ZhangXiaohan QinYu ChengJunchi Yan
Topic ModelingNatural Language Processing TechniquesAdvanced Neural Network ApplicationsXiangdong Zhang 1 2 Debing Zhang 2 Shaofeng Zhang 3 Xiaohan Qin 1 2 Yu Cheng 4 Junchi Yan 1
PublishedMay 24, 2026
TypePreprint
Citations0

Powered by the Exa API