ウェブファーマー List of News Articles

We held the "2nd World Model and Its Surroundings" online, hosted by our company. This time, Mr. Ishii presented the latest paper by LeCun, the proponent of the world model, and our representative, Mr. Omasa, presented the DiWA model. After the presentations, various questions were raised by the participants, leading to discussions on the architecture of the world model, the latest trends, and the potential applications of the world model in 3D representation and future predictions.

On February 16, 2026, we held the "4th LLM Relearning Forum" online. This time, Mr. Nakamura from JIN Corporation and Representative Yamaguchi from Bankom Corporation also presented, and we discussed topics such as OpenClaw and RLHF.

On January 27, 2026, we held the "1st World Model and its Surroundings" online. This study session aims to deepen understanding of World Model, Physical AI, and Vision Foundation Model. In this first session, after a general overview of world models and Physical AI, a presentation was made by the representative, Omasa, on the VGGT model, NWM model, and what follows. Many people tuned in to watch.

On December 11, 2025, we held the "3rd LLM Relearning Forum" online. The purpose of this study session is to deepen our understanding of LLM relearning. This time, we discussed the relearning methods of GPT OSS and other LLM-related topics. Once again, we had Mr. Yamaguchi from Bankom Co., Ltd. present, touching on LLM relearning and general topics related to LLMs.

The "2nd LLM Relearning Symposium" was held online on October 26, 2025, organized by WEBFARMER. The purpose of this study session is to deepen understanding of LLM relearning. This time, we will discuss the relearning methods of GPT OSS that were announced in August. We also had a presentation from Mr. Yamaguchi, the representative of Bankom Co., Ltd., who touched on topics related to LLM relearning and LLMs in general.

On September 24, 2025, the first online seminar on the topic of LLM retraining was held. The purpose of this study session is to deepen understanding of LLM retraining. This time, we focused on chapters 6 and 7 of "Introduction to Creating LLMs" (by Mynavi) and discussed methods of retraining. As a Webfarmer who has been involved in computer vision studies for a long time, I shared insights on key points regarding LLM retraining from that perspective. In future sessions, we plan to widely invite participants and touch on various topics related to LLM retraining.