EXAONE Deep: Reasoning Enhanced Language Models
This work addresses the need for more effective reasoning in AI models, particularly for tasks like math and coding, though it appears incremental as it builds on existing language model paradigms with specialized data.
The paper tackles the problem of enhancing reasoning capabilities in language models by training on reasoning-specialized datasets with long thought processes, resulting in smaller models outperforming comparable-sized ones and the largest model achieving competitive performance against leading open-weight models.
We present EXAONE Deep series, which exhibits superior capabilities in various reasoning tasks, including math and coding benchmarks. We train our models mainly on the reasoning-specialized dataset that incorporates long streams of thought processes. Evaluation results show that our smaller models, EXAONE Deep 2.4B and 7.8B, outperform other models of comparable size, while the largest model, EXAONE Deep 32B, demonstrates competitive performance against leading open-weight models. All EXAONE Deep models are openly available for research purposes and can be downloaded from https://huggingface.co/LGAI-EXAONE