LG AI Research Unveils EXAONE Deep: A Reasoning Model for Complex Problem-Solving
A Leap Forward in Artificial Intelligence
LG AI Research has made a significant breakthrough in the development of advanced reasoning models, introducing EXAONE Deep, a model that excels in complex problem-solving across mathematics, science, and coding. This achievement is a major milestone in the realm of artificial intelligence, as only a handful of organizations have successfully created foundational models for complex reasoning.
Exceptional Reasoning Capabilities
The EXAONE Deep model has demonstrated exceptional reasoning abilities in core domains, showcasing a strong ability to understand and apply knowledge across a broader range of subjects. The model’s performance benchmarks are impressive, outperforming competing models in various areas.
Mathematics
The 32B model outperformed a competing model, despite being only 5% of its size, in a demanding mathematics benchmark. Additionally, the 7.8B and 2.4B versions achieved top rankings in all major mathematics benchmarks for their respective model sizes.
Science and Coding
In science and coding, the EXAONE Deep models (7.8B and 2.4B) secured the top spot across all major benchmarks.
MMLU (Massive Multitask Language Understanding)
The 32B model achieved a score of 83.0 on the MMLU benchmark, the best performance among domestic Korean models.
Global Recognition
The capabilities of the EXAONE Deep 32B model have already garnered international recognition. Shortly after its release, it was included in the ‘Notable AI Models’ list by US-based non-profit research organization Epoch AI, making LG the only Korean entity with models featured on this prestigious list in the past two years.
Maths Prowess
EXAONE Deep has demonstrated exceptional mathematical reasoning skills across its various model sizes (32B, 7.8B, and 2.4B). In assessments based on the 2025 academic year’s mathematics curriculum, all three models outperformed global reasoning models of comparable size.
Science and Coding Excellence
EXAONE Deep has showcased remarkable capabilities in professional science reasoning and software coding. The 32B model scored 66.1 on the GPQA Diamond test, which assesses problem-solving skills in doctoral-level physics, chemistry, and biology. In the LiveCodeBench evaluation, which measures coding proficiency, the model achieved a score of 59.5, indicating its potential for high-level applications in these expert domains.
Enhanced General Knowledge
Beyond its specialized reasoning capabilities, EXAONE Deep has also demonstrated improved performance in general knowledge understanding. The 32B model achieved an impressive score of 83.0 on the MMLU benchmark, positioning it as the top-performing domestic model in this comprehensive evaluation.
Conclusion
LG AI Research believes that EXAONE Deep’s reasoning advancements represent a leap towards a future where AI can tackle increasingly complex problems and contribute to enriching and simplifying human lives through continuous research and innovation.
Frequently Asked Questions
Q: What is the main achievement of EXAONE Deep?
A: EXAONE Deep is a reasoning model that excels in complex problem-solving across mathematics, science, and coding.
Q: What are the key strengths of EXAONE Deep?
A: EXAONE Deep has demonstrated exceptional reasoning abilities in core domains, showcasing a strong ability to understand and apply knowledge across a broader range of subjects.
Q: How does EXAONE Deep perform in mathematics, science, and coding?
A: EXAONE Deep has outperformed competing models in various areas, including mathematics, science, and coding, securing top rankings in major benchmarks.
Q: What is the significance of EXAONE Deep’s inclusion in the ‘Notable AI Models’ list?
A: EXAONE Deep’s inclusion in the ‘Notable AI Models’ list by Epoch AI highlights its international recognition and puts LG AI Research among the top players in the AI research community.
