DecisionLLM: Large Language Models for Long Sequence Decision Exploration
Fuente:
arXiv
Enregistré dans:
| Auteurs principaux: | , , , , , , , , , , , , |
|---|---|
| Format: | Preprint |
| Publié: |
2026
|
| Sujets: | |
| Accès en ligne: | |
| Tags: |
Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
|
| _version_ | 1866917205146337280 |
|---|---|
| author | Lv, Xiaowei Zhang, Zhilin Li, Yijun Huo, Yusen Ju, Siyuan Li, Xuyan Hong, Chunxiang Wang, Tianyu Wang, Yongcai Sun, Peng Yu, Chuan Xu, Jian Zheng, Bo |
| author_facet | Lv, Xiaowei Zhang, Zhilin Li, Yijun Huo, Yusen Ju, Siyuan Li, Xuyan Hong, Chunxiang Wang, Tianyu Wang, Yongcai Sun, Peng Yu, Chuan Xu, Jian Zheng, Bo |
| contents | Long-sequence decision-making, which is usually addressed through reinforcement learning (RL), is a critical component for optimizing strategic operations in dynamic environments, such as real-time bidding in computational advertising. The Decision Transformer (DT) introduced a powerful paradigm by framing RL as an autoregressive sequence modeling problem. Concurrently, Large Language Models (LLMs) have demonstrated remarkable success in complex reasoning and planning tasks. This inspires us whether LLMs, which share the same Transformer foundation, but operate at a much larger scale, can unlock new levels of performance in long-horizon sequential decision-making problem. This work investigates the application of LLMs to offline decision making tasks. A fundamental challenge in this domain is the LLMs' inherent inability to interpret continuous values, as they lack a native understanding of numerical magnitude and order when values are represented as text strings. To address this, we propose treating trajectories as a distinct modality. By learning to align trajectory data with natural language task descriptions, our model can autoregressively predict future decisions within a cohesive framework we term DecisionLLM. We establish a set of scaling laws governing this paradigm, demonstrating that performance hinges on three factors: model scale, data volume, and data quality. In offline experimental benchmarks and bidding scenarios, DecisionLLM achieves strong performance. Specifically, DecisionLLM-3B outperforms the traditional Decision Transformer (DT) by 69.4 on Maze2D umaze-v1 and by 0.085 on AuctionNet. It extends the AIGB paradigm and points to promising directions for future exploration in online bidding. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2601_10148 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | DecisionLLM: Large Language Models for Long Sequence Decision Exploration Lv, Xiaowei Zhang, Zhilin Li, Yijun Huo, Yusen Ju, Siyuan Li, Xuyan Hong, Chunxiang Wang, Tianyu Wang, Yongcai Sun, Peng Yu, Chuan Xu, Jian Zheng, Bo Artificial Intelligence Long-sequence decision-making, which is usually addressed through reinforcement learning (RL), is a critical component for optimizing strategic operations in dynamic environments, such as real-time bidding in computational advertising. The Decision Transformer (DT) introduced a powerful paradigm by framing RL as an autoregressive sequence modeling problem. Concurrently, Large Language Models (LLMs) have demonstrated remarkable success in complex reasoning and planning tasks. This inspires us whether LLMs, which share the same Transformer foundation, but operate at a much larger scale, can unlock new levels of performance in long-horizon sequential decision-making problem. This work investigates the application of LLMs to offline decision making tasks. A fundamental challenge in this domain is the LLMs' inherent inability to interpret continuous values, as they lack a native understanding of numerical magnitude and order when values are represented as text strings. To address this, we propose treating trajectories as a distinct modality. By learning to align trajectory data with natural language task descriptions, our model can autoregressively predict future decisions within a cohesive framework we term DecisionLLM. We establish a set of scaling laws governing this paradigm, demonstrating that performance hinges on three factors: model scale, data volume, and data quality. In offline experimental benchmarks and bidding scenarios, DecisionLLM achieves strong performance. Specifically, DecisionLLM-3B outperforms the traditional Decision Transformer (DT) by 69.4 on Maze2D umaze-v1 and by 0.085 on AuctionNet. It extends the AIGB paradigm and points to promising directions for future exploration in online bidding. |
| title | DecisionLLM: Large Language Models for Long Sequence Decision Exploration |
| topic | Artificial Intelligence |
| url | https://arxiv.org/abs/2601.10148 |