TimeBraid, a series of unified time-series and language models that align pretrained language models and pretrained time-series foundation models through interleaved global residual attention layers, remains competitive with far larger general-purpose models and task-specific counterparts.
Abstract
We present TimeBraid, a series of unified time-series and language models that align pretrained language models and pretrained time-series foundation models through interleaved global residual attention layers. Each model inherits knowledge, instruction following, and reasoning from one side, continuous-signal perception and zero-shot forecasting from the other, and fuses the two in a shared representation space where both modalities are understood and generated. We study the design choices that make such unified modeling work: where to align the two representation spaces, how to ground language in temporal structure, how to balance understanding with generation, and how to keep joint optimization stable. The resulting recipe combines a unified prompting scheme for diverse time-series and text tasks, stabilized joint training, and supervision from 2.2M curated series--text pairs and 4.9M instruction-tuning samples. Across benchmarks spanning time-series perception, understanding, reasoning, and both context-aided and unimodal forecasting, TimeBraid remains competitive with far larger general-purpose models and task-specific counterparts.
OpenTSLM TeeMoE is introduced, a generalist time-series language model that can forecast directly from observed time series, reason over textual context and temporal patterns, and synthesize and refine predictions from external numerical forecasting specialists.
Tony Chen, Timo Stoffregen, Maxwell A. Xu et al.· arXiv.org· 0 citations
Results indicate that freezing the pretrained backbone and training lightweight projection and adapter modules provides a favourable accuracy--efficiency trade-off with stable behaviour across varying context lengths.
Tamanna S. Kumavat, Georg Brunner, Kyriakos Flouris· arXiv.org· 0 citations
CGTime, the 4B-parameter computation-grounded time-series-language model, decoupling perception from description, outperforms far larger general-purpose models on multivariate understanding tasks and attains the best multivariate fact score on a held-out benchmark.
Xinran Feng, Yi Xie, Chao Zhang et al.· 1 citation· ⚡1
This work develops a unified taxonomy of memory mechanisms and reviews three classes of external memory, including explicit modules, retrieval augmentation, and agentic stores, under a common framework for what is retained, how it is written and accessed, and how it persists.
M. Nguyen, H. Nguyen, Manh Nguyen et al.· 0 citations
Tabby, a long context probabilistic time series foundation model, is released together with a complete and open recipe of how it was built, which achieves competitive zero-shot forecasting performance on GIFT-Eval and the out-of-distribution TIME benchmark.
Shi-Feng Xie, Bahaeddine Abdessalem, Ze-Hao Xiao et al.· 1 citation
This work proposes SwitchPFN, which learns a shared projection and regime codebook from the training sequences, making local dynamic operators and transition features directly comparable across samples, and achieves the highest mean accuracy among the evaluated methods.
Zhen-Yi Zhu, Jacqueline J. Pang, Pei-Lin Shen et al.· arXiv.org· 0 citations