6 papers
JUBAKU: An Adversarial Benchmark for Exposing Culturally Grounded Stereotypes in Japanese LLMs
Taihei Shiotani, Masahiro Kaneko, Ayana Niwa +4
Social biases reflected in language are inherently shaped by cultural norms, which vary significantly across regions and lead to diverse manifestations of stereotypes. Existing eva…
Multi-modal, Multi-task, Multi-criteria Automatic Evaluation with Vision Language Models
Masanari Ohi, Masahiro Kaneko, Naoaki Okazaki +1
Vision-language models (VLMs) have shown impressive abilities across a range of multi-modal tasks. However, existing metrics for evaluating the quality of text generated by VLMs ty…
Rewriting Pre-Training Data Boosts LLM Performance in Math and Code
Kazuki Fujii, Yukito Tajima, Sakae Mizuki +14
The performance of large language models (LLMs) in program synthesis and mathematical reasoning is fundamentally limited by the quality of their pre-training corpora. We introduce…
Likelihood-based Mitigation of Evaluation Bias in Large Language Models
Masanari Oi, Masahiro Kaneko, Ryuto Koike +2
Large Language Models (LLMs) are widely used to evaluate natural language generation tasks as automated metrics. However, the likelihood, a measure of LLM's plausibility for a sent…
Why We Build Local Large Language Models: An Observational Analysis from 35 Japanese and Multilingual LLMs
Koshiro Saito, Sakae Mizuki, Masanari Ohi +11
Why do we build local large language models (LLMs)? What should a local LLM learn from the target language? Which abilities can be transferred from other languages? Do language-spe…
Building Instruction-Tuning Datasets from Human-Written Instructions with Open-Weight Large Language Models
Youmi Ma, Sakae Mizuki, Kazuki Fujii +12
Instruction tuning is crucial for enabling Large Language Models (LLMs) to solve real-world tasks. Prior work has shown the effectiveness of instruction-tuning data synthesized sol…