Papers by Kuniko Saito
Let’s Put Ourselves in Sally’s Shoes: Shoes-of-Others Prefilling Improves Theory of Mind in Large Language Models (2026.findings-eacl)
Copied to clipboard
Kazutoshi Shinoda, Nobukatsu Hojo, Kyosuke Nishida, Yoshihiro Yamazaki, Keita Suzuki, Hiroaki Sugiyama, Kuniko Saito
| Challenge: | Existing methods for Theory of Mind (ToM) are specialized for inferring beliefs from contexts involving changes in the world state. |
| Approach: | They propose a method which makes fewer assumptions about contexts and is applicable to broader scenarios. |
| Outcome: | The proposed method makes fewer assumptions about contexts and is applicable to broader scenarios. |
Initialization of Large Language Models via Reparameterization to Mitigate Loss Spikes (2024.emnlp-main)
Copied to clipboard
| Challenge: | Existing methods to train large language models that require a non-uniform model norm are not effective. |
| Approach: | They propose a technique that allows for uniformity of the norm of the model parameters . they propose 'weight scaling as reparameterization' to adjust the norm to the parameter . |
| Outcome: | The proposed technique outperforms existing methods and stabilizes training with the transformer decoders. |
Combining Argumentation Structure and Language Model for Generating Natural Argumentative Dialogue (2022.aacl-short)
Copied to clipboard
| Challenge: | Argumentative dialogue is important process where speakers discuss a specific theme for consensus building or decision making. |
| Approach: | They propose a method to generate argumentative dialogues by combining argumentation structure and language model. |
| Outcome: | The proposed method significantly improves the naturalness of arguments without losing consistency. |
DueT: Image-Text Contrastive Transfer Learning with Dual-adapter Tuning (2023.emnlp-main)
Copied to clipboard
| Challenge: | Comparative learning models for vision and language models are gaining popularity . dueT trains only adapters inserted into pre-trained image and text encoders . |
| Approach: | They propose a transfer learning method for vision and language models built by contrastive learning that trains only adapters inserted into the frozen image and text encoders. |
| Outcome: | The proposed method outperforms fine-tuning, and the LoRA-based adapter method in English and Japanese domains. |