TIV: Thought Injection via Vectors for Efficient Reasoning in Large Reasoning Models
Yi Cao, Weijie Shi, Wei-Jie Xu, Yucheng Shen, Yue Cui, Hanghui Guo, Shimin Di, Ziyi Liu
Abstract
Large Reasoning Models (LRMs) have recently demonstrated impressive performance across a range of reasoning tasks by generating intermediate thoughts. However, these models can suffer from overthinking—generating excessive tokens that contribute little to final accuracy while increasing inference cost. To mitigate this, we propose TIV (Thought Injection via Vectors), an innovative framework that compresses token-level reasoning into compact vectors without sacrificing performance. Rather than generating explicit thoughts, TIV injects learnable vectors into the post-attention hidden states of the final token across Transformer layers, enabling implicit and lightweight reasoning. We further introduce a two-stage reinforcement learning strategy: the first stage calibrates the model
BibTeX
@inproceedings{aaai2026_tivthoughtinject,
title = {TIV: Thought Injection via Vectors for Efficient Reasoning in Large Reasoning Models},
author = {Yi Cao and Weijie Shi and Wei-Jie Xu and Yucheng Shen and Yue Cui and Hanghui Guo and Shimin Di and Ziyi Liu and Jiaming Li and Alexander Zhou and Jia Zhu and Jiajie Xu},
booktitle = {AAAI 2026},
year = {2026}
}