Everyday Physics in Korean Contexts: A Culturally Grounded Physical Reasoning Benchmark
On the map: LLM reasoning → Reasoning benchmarks and evaluation
Abstract (arXiv)
Existing physical commonsense reasoning benchmarks predominantly focus on Western contexts, overlooking cultural variations in physical problem-solving. To address this gap, we introduce EPiK (Everyday Physics in Korean Contexts), a novel benchmark comprising 181 binary-choice problems that test physical reasoning within Korean cultural contexts, ranging from kimchi (Korean food) to traditional fermentation. EPiK is constructed using a two-stage generation and verification pipeline to create culturally-authentic problems across 9 reasoning subtasks and 84 scenarios. Unlike approaches based on simple translation, our method generates problems organically from Korean contexts while upholding rigorous physical reasoning standards. Our evaluations show that Korean-specialized models consistently outperform general-purpose models of comparable size. This performance gap highlights the limitations of culturally-agnostic models and demonstrates the critical need for culturally-aware benchmarks to truly measure language understanding. Our EPiK is publicly available at https://huggingface.co/datasets/jjae/EPiK.
Key facts. Cited by 1 paper in the map (data through 14 Oct 2026). Placed in Reasoning benchmarks and evaluation, one of 17 directions in LLM reasoning. Builds on 25 of its 40 references that are in the map. First posted on arXiv 22 Sep 2025.
Builds on
25 of 40 references.
Closest ancestors · 4
Same direction, newest first
- 2025EXAONE 4.0: Unified Large Language Models Integrating Non-reasoning and Reasoning Modes15 Jul
- 2024mCSQA: Multilingual Commonsense Reasoning Dataset with Unified Creation Strategy by Language Models and Humans6 Jun
- 2024IndoCulture: Exploring Geographically-Influenced Cultural Commonsense Reasoning Across Eleven Indonesian Provinces2 Apr
- 2023Benchmarks for Automated Commonsense Reasoning: A Survey9 Feb
Other references · 21
- 2024The Llama 3 Herd of Modelscited by 14,715
- 2023Efficient Memory Management for Large Language Model Serving with PagedAttentioncited by 5,944
- 2019PIQA: Reasoning about Physical Commonsense in Natural Languagecited by 2,883
- 2024Gemma 2: Improving Open Language Models at a Practical Sizecited by 1,998
- 2020XCOPA: A Multilingual Dataset for Causal Commonsense Reasoningcited by 333
- 2022Challenges and Strategies in Cross-Cultural NLPcited by 218
- 2024BLEnD: A Benchmark for LLMs on Everyday Knowledge in Diverse Cultures and Languagescited by 173
- 2024CultureBank: An Online Community-Driven Knowledge Base Towards Culturally Aware Language Technologiescited by 98
- 2020Translation Artifacts in Cross-lingual Transfer Learningcited by 96
- 2024CLIcK: A Benchmark Dataset of Cultural and Linguistic Intelligence in Koreancited by 66
- 2021Common Sense Beyond English: Evaluating and Improving Multilingual Language Models for Commonsense Reasoningcited by 65
- 2020Temporal Reasoning on Implicit Events from Distant Supervisioncited by 58
9 more
- 2024EXAONE 3.5: Series of Large Language Models for Real-world Use Casescited by 53
- 2024Understanding the Capabilities and Limitations of Large Language Models for Cultural Commonsensecited by 51
- 2023NEWTON: Are Large Language Models Capable of Physical Reasoning?cited by 45
- 2023Revisiting Machine Translation for Cross-lingual Classificationcited by 41
- 2021The World of an Octopus: How Reporting Bias Influences a Language Model's Perception of Colorcited by 32
- 2020Towards an Atlas of Cultural Commonsense for Machine Reasoningcited by 29
- 2025Kanana: Compute-efficient Bilingual Language Modelscited by 20
- 2024Do Multilingual Large Language Models Mitigate Stereotype Bias?cited by 13
- 2024Susu Box or Piggy Bank: Assessing Cultural Commonsense Knowledge between Ghana and the U.Scited by 9
Built on by
Citing papers per year: 2026: 1.
1 paper.
Similar, not cited
Close in wording, no citation either way
- 2025Ko-PIQA: A Korean Physical Commonsense Reasoning Dataset with Cultural Context
- 2025Nunchi-Bench: Benchmarking Language Models on Cultural Reasoning with a Focus on Korean Superstition
- 2025Multi-Step Reasoning in Korean and the Emergent Mirage
- 2025Understand, Solve and Translate: Bridging the Multilingual Mathematical Reasoning Gap
- 2025ABench-Physics: Benchmarking Physical Reasoning in LLMs via High-Difficulty and Dynamic Physics Problems
- 2025KoSimpleQA: A Korean Factuality Benchmark with an Analysis of Reasoning LLMs
- 2025Pushing on Multilingual Reasoning Models with Language-Mixed Chain-of-Thought
- 2025PhysReason: A Comprehensive Benchmark towards Physics-Based Reasoning