VAIOM 是专门处理金融连续数据的解码器模型，在汇率数据上比 LightGBM 强 0.03 比特以上，创新点是把连续输入和离散输出分开建模，值得做时序预测的人一看。
VAIOM 是一种 decoder-only Transformer，专门用于对一小时的汇率柱进行概率性下一回报建模。它采用连续多元金融事件向量作为输入，同时将输出建模为波动率归一化后收益桶上的类别分布。在 0.9M 混合连续输入模型上，结合了连续事件特征和分类资产元数据，以及混合市场状态回报头、间隙、波动率制度和序数辅助目标，并通过全序列监督进行训练。在 2025 年的两个测试周期中，VAIOM 在所有训练种子上均优于固定单柱 LightGBM 基线，标准检查点的增益分别为每个事件 0.029 和 0.043 比特。对比实验表明，连续输入优于离散令牌输入，全序列监督优于最后一个位置训练，且辅助表征塑造和混合结构回报头改善了回报似然。
VAIOM: Continuous-Input, Discrete-Output Decoder-Only Financial Sequence Modeling
Financial observations are continuous, heterogeneous, and noisy, whereas decoder-only next-token models are usually built around discrete symbolic inputs. We introduce Vector-Input Autoregressive Inference for Ordinal-Return Modeling (VAIOM), a decoder-only Transformer for probabilistic next-return modeling on one-hour foreign-exchange bars. VAIOM separates input representation from output likelihood: continuous multivariate financial-event vectors preserve numerical structure at the input, while a categorical distribution over the next volatility-normalized return bucket supports cross-entropy training and likelihood evaluation. The selected 0.9M Hybrid Continuous Input model combines continuous event features with categorical asset metadata, a Mixture-of-Market-States return head, Gap, volatility-regime, and Ordinal auxiliary objectives, and full-sequence supervision. Models and preprocessing are fit using pre-2024 Train data; models are selected on 2024H2 Validation and evaluated without refitting on two 2025 Test periods. Across three independent training seeds, every model outperforms fixed single-bar LightGBM baseline in both Test halves. For the canonical checkpoint, paired gains over LightGBM are 0.029 and 0.043 bits per event. Validation experiments show that continuous input improves over discrete-token input under the same categorical return objective, full-sequence supervision improves over last-position training, and auxiliary representation shaping together with a mixture-structured return head improves return likelihood in controlled comparisons. A supporting capacity study finds that the smallest evaluated complete architecture rung achieves the strongest Validation likelihood on the present corpus.