Retrieva to Exhibit LLM Vulnerability Diagnostic Web Application "KOKOROE" at the "GENIAC-PRIZE Results Presentation Caravan"
Retrieva Inc. announces that it will exhibit "KOKOROE", an LLM vulnerability diagnostic web application co-developed with Everyleaf, at the NEDO-sponsored "GENIAC-PRIZE Results Presentation Caravan" in Shibuya on April 14, 2026.
📋 Article Processing Timeline
- 📰 Published: March 31, 2026 at 20:00
- 🔍 Collected: April 1, 2026 at 13:39 (17h 39m after Published)
- 🤖 AI Analyzed: April 22, 2026 at 03:29 (493h 50m after Collected)
Retrieva Inc. (Toshima-ku, Tokyo, CEO: Takuya Taguchi) announces that it will exhibit the LLM vulnerability diagnostic web application "KOKOROE" at the NEDO prize-money utilization program "GENIAC-PRIZE Results Presentation Caravan" to be held at Shibuya Stream Hall, Shibuya-ku, Tokyo on April 14, 2026.
## About the Exhibition Content
We will exhibit "KOKOROE," which our company co-developed with our group company Everyleaf Corporation.
"KOKOROE" is a web application that can diagnose whether an LLM is providing safe answers and improve the LLM based on the diagnostic results. By humans simply providing very plain guidelines, the LLM can be tuned so that it can be safely utilized according to the purpose of use.
In recent years, the introduction and utilization of Generative AI have progressed in many companies against the backdrop of rising expectations for operational improvement and productivity enhancement, but LLM responses are not always safe. In addition to responses that promote crimes, there are risks specific to each industry, such as promoting dangerous work procedures in the manufacturing industry and making incorrect judgments regarding medicinal efficacy in the pharmaceutical industry.
"KOKOROE," which we are exhibiting this time, responds to the risks and vulnerabilities of various industries through a mechanism that automatically diagnoses and expands using very plain guidelines written by humans.
## About the Background of the Exhibition
Our company has been working on research and development with an emphasis on securing "safety" in addition to "usefulness" so that companies can safely and effectively utilize Large Language Models (LLMs) in their business operations.
Against this background, our company participated in the "Development of Risk Exploration and Risk Reduction Technologies for Securing the Safety of Generative AI" in the GENIAC PRIZE, and has promoted the visualization of potential risks inherent in Generative AI and the development of technologies contributing to their reduction.
At this caravan, by broadly sharing the results of these efforts, we aim to promote discussions on improving the safety of Generative AI and contribute to the improvement of safety in society as a whole.
## About the GENIAC-PRIZE Results Presentation Caravan
The "GENIAC-PRIZE" is a prize-money type contest led by the Ministry of Economy, Trade and Industry and the New Energy and Industrial Technology Development Organization (NEDO), aimed at accelerating the social implementation of Generative AI.
It sets four themes expected to be solved by Generative AI, and recruits and examines the development of applications based on specific needs.
4 themes expected to be solved by Generative AI:
- Formalizing tacit knowledge in the manufacturing industry
- Improving productivity in customer support
- Streamlining screening operations in government offices
- Developing risk exploration and risk reduction technologies to ensure the safety of Generative AI
At the results presentation caravan held as part of this initiative, in addition to the introduction of Generative AI use cases by award winners and applicants, opportunities for interaction and networking with participants will be provided.
## Caravan Overview
GENIAC-PRIZE Results Presentation Caravan
Date: Tuesday, April 14, 2026 10:00-17:00
Venue: Shibuya Stream Hall
Organizer: Ministry of Economy, Trade and Industry, Commerce and Information Policy Bureau
## About the Exhibition Content
We will exhibit "KOKOROE," which our company co-developed with our group company Everyleaf Corporation.
"KOKOROE" is a web application that can diagnose whether an LLM is providing safe answers and improve the LLM based on the diagnostic results. By humans simply providing very plain guidelines, the LLM can be tuned so that it can be safely utilized according to the purpose of use.
In recent years, the introduction and utilization of Generative AI have progressed in many companies against the backdrop of rising expectations for operational improvement and productivity enhancement, but LLM responses are not always safe. In addition to responses that promote crimes, there are risks specific to each industry, such as promoting dangerous work procedures in the manufacturing industry and making incorrect judgments regarding medicinal efficacy in the pharmaceutical industry.
"KOKOROE," which we are exhibiting this time, responds to the risks and vulnerabilities of various industries through a mechanism that automatically diagnoses and expands using very plain guidelines written by humans.
## About the Background of the Exhibition
Our company has been working on research and development with an emphasis on securing "safety" in addition to "usefulness" so that companies can safely and effectively utilize Large Language Models (LLMs) in their business operations.
Against this background, our company participated in the "Development of Risk Exploration and Risk Reduction Technologies for Securing the Safety of Generative AI" in the GENIAC PRIZE, and has promoted the visualization of potential risks inherent in Generative AI and the development of technologies contributing to their reduction.
At this caravan, by broadly sharing the results of these efforts, we aim to promote discussions on improving the safety of Generative AI and contribute to the improvement of safety in society as a whole.
## About the GENIAC-PRIZE Results Presentation Caravan
The "GENIAC-PRIZE" is a prize-money type contest led by the Ministry of Economy, Trade and Industry and the New Energy and Industrial Technology Development Organization (NEDO), aimed at accelerating the social implementation of Generative AI.
It sets four themes expected to be solved by Generative AI, and recruits and examines the development of applications based on specific needs.
4 themes expected to be solved by Generative AI:
- Formalizing tacit knowledge in the manufacturing industry
- Improving productivity in customer support
- Streamlining screening operations in government offices
- Developing risk exploration and risk reduction technologies to ensure the safety of Generative AI
At the results presentation caravan held as part of this initiative, in addition to the introduction of Generative AI use cases by award winners and applicants, opportunities for interaction and networking with participants will be provided.
## Caravan Overview
GENIAC-PRIZE Results Presentation Caravan
Date: Tuesday, April 14, 2026 10:00-17:00
Venue: Shibuya Stream Hall
Organizer: Ministry of Economy, Trade and Industry, Commerce and Information Policy Bureau