# NOLLI: A Difficulty-Calibrated Puzzle Benchmark for Diagnosing the English-Korean Performance Gap

arXiv: https://arxiv.org/abs/2608.04397 · HTML: https://mlsanity.com/paper/2608.04397/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 27 of its 38 references that are in the map.
- First posted on arXiv 5 Aug 2026.

## Builds on (most cited first, 27 of 27)

- [Language Models are Few-Shot Learners](https://mlsanity.com/paper/2005.14165/) (2020)
- [Training Verifiers to Solve Math Word Problems](https://mlsanity.com/paper/2110.14168/) (2021)
- [Measuring Mathematical Problem Solving With the MATH Dataset](https://mlsanity.com/paper/2103.03874/) (2021)
- [DeepSeek-V4: Towards Highly Efficient Million-Token Context Intelligence](https://mlsanity.com/paper/2606.19348/) (2026)
- [Language Models are Multilingual Chain-of-Thought Reasoners](https://mlsanity.com/paper/2210.03057/) (2022)
- [ByT5: Towards a token-free future with pre-trained byte-to-byte models](https://mlsanity.com/paper/2105.13626/) (2021)
- [On the Planning Abilities of Large Language Models : A Critical Investigation](https://mlsanity.com/paper/2305.15771/) (2023)
- [ZebraLogic: On the Scaling Limits of LLMs for Logical Reasoning](https://mlsanity.com/paper/2502.01100/) (2025)
- [A Careful Examination of Large Language Model Performance on Grade School Arithmetic](https://mlsanity.com/paper/2405.00332/) (2024)
- [CLUTRR: A Diagnostic Benchmark for Inductive Reasoning from Text](https://mlsanity.com/paper/1908.06177/) (2019)
- [Stop Uploading Test Data in Plain Text: Practical Strategies for Mitigating Data Contamination by Evaluation Benchmarks](https://mlsanity.com/paper/2305.10160/) (2023)
- [DyVal: Dynamic Evaluation of Large Language Models for Reasoning Tasks](https://mlsanity.com/paper/2309.17167/) (2023)
- [KMMLU: Measuring Massive Multitask Language Understanding in Korean](https://mlsanity.com/paper/2402.11548/) (2024)
- [NPHardEval: Dynamic Benchmark on Reasoning Ability of Large Language Models via Complexity Classes](https://mlsanity.com/paper/2312.14890/) (2023)
- [CLIcK: A Benchmark Dataset of Cultural and Linguistic Intelligence in Korean](https://mlsanity.com/paper/2403.06412/) (2024)
- [Comparing Test Sets with Item Response Theory](https://mlsanity.com/paper/2106.00840/) (2021)
- [An Empirical Study of Tokenization Strategies for Various Korean NLP Tasks](https://mlsanity.com/paper/2010.02534/) (2020)
- [SATBench: Benchmarking LLMs' Logical Reasoning via Automated Puzzle Generation from SAT Formulas](https://mlsanity.com/paper/2505.14615/) (2025)
- [EXAONE 4.0: Unified Large Language Models Integrating Non-reasoning and Reasoning Modes](https://mlsanity.com/paper/2507.11407/) (2025)
- [CUTE: Measuring LLMs' Understanding of Their Tokens](https://mlsanity.com/paper/2409.15452/) (2024)
- [AutoLogi: Automated Generation of Logic Puzzles for Evaluating Reasoning Abilities of Large Language Models](https://mlsanity.com/paper/2502.16906/) (2025)
- [Mi:dm 2.0 Korea-centric Bilingual Language Models](https://mlsanity.com/paper/2601.09066/) (2026)
- [Multilingual Reasoning Gym: Multilingual Scaling of Procedural Reasoning Environments](https://mlsanity.com/paper/2603.10793/) (2026)
- [Multi-Step Reasoning in Korean and the Emergent Mirage](https://mlsanity.com/paper/2501.05712/) (2025)
- [EXECUTE: A Multilingual Benchmark for LLM Token Understanding](https://mlsanity.com/paper/2505.17784/) (2025)
- [Enhancing Character-Level Understanding in LLMs through Token Internal Structure Learning](https://mlsanity.com/paper/2411.17679/) (2024)
- [Redefining Evaluation Standards: A Unified Framework for Evaluating the Korean Capabilities of Language Models](https://mlsanity.com/paper/2503.22968/) (2025)

## Abstract (arXiv)

We introduce NOLLI, a procedurally generated English-Korean puzzle benchmark designed to diagnose where Korean performance gaps arise. It comprises 15 puzzle types (25 tasks; 7,500 items), with every instance seed-regenerable, verified to have a unique solution, and scored deterministically. Rather than equating harder with bigger, we calibrate difficulty behaviorally, tuning each generator until a fixed reference model lands in target accuracy bands. Its three-level design combines matched direct translations, script adaptations over Hangul jamo (sub-syllabic letters), and Korean-only tasks grounded in Korean culture or orthography. We evaluate 15 frontier, open-weight, and Korean-developed models; among the 12 above a 3% overall-accuracy floor, matched English-Korean accuracy is statistically equivalent within a +/- 10 pp margin (TOST), suggesting little cost from presentation language alone. Writing-system-intensive tasks show sharper gaps: Korean Cipher falls behind English by up to 68.7 pp, whereas Cryptarithmetic over the same jamo shows no systematic penalty, and Jamo Composition accuracy predicts Korean Cipher accuracy. These contrasts are diagnostic rather than causal, consistent with difficulty in multi-step sub-syllabic execution. Korean-only tasks separate rule-application deficits, which vary in sign, from a Kinship deficit positive in all 12. Finally, a salient size measure fails to grow from Easy to Hard in 7 of 15 types, making structural size an unreliable proxy for empirical difficulty.