Search papers, labs, and topics across Lattice.
This paper introduces Large Causal Models (LCMs), a novel class of pre-trained neural architectures for temporal causal discovery, addressing the limitations of dataset-specific models and the lack of multi-dataset pretraining. They train LCMs using a combination of diverse synthetic generators and realistic time-series datasets, enabling scalability to higher variable counts and deeper architectures. Experimental results on synthetic, semi-synthetic, and realistic benchmarks demonstrate that LCMs achieve competitive or superior accuracy compared to classical and neural baselines, especially in out-of-distribution settings, while also enabling fast, single-pass inference.
Forget dataset-specific models: Large Causal Models (LCMs) pre-trained on diverse data now offer fast, scalable, and accurate temporal causal discovery, even outperforming baselines in out-of-distribution scenarios.
Causal discovery for both cross-sectional and temporal data has traditionally followed a dataset-specific paradigm, where a new model is fitted for each individual dataset. Such an approach limits the potential of multi-dataset pretraining. The concept of large causal models (LCMs) envisions a class of pre-trained neural architectures specifically designed for temporal causal discovery. Prior approaches are constrained to small variable counts, degrade with larger inputs, and rely heavily on synthetic data, limiting generalization. We propose a principled framework for LCMs, combining diverse synthetic generators with realistic time-series datasets, allowing learning at scale. Extensive experiments on synthetic, semi-synthetic and realistic benchmarks show that LCMs scale effectively to higher variable counts and deeper architectures while maintaining strong performance. Trained models achieve competitive or superior accuracy compared to classical and neural baselines, particularly in out-of-distribution settings, while enabling fast, single-pass inference. Results demonstrate LCMs as a promising foundation-model paradigm for temporal causal discovery. Experiments and model weights are available at https://github.com/kougioulis/LCM-paper/.
