Ana Parra-Astudillo
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Ana Parra-Astudillo has a total of 19 co-authors. Here are the top co-authors by number of shared publications:
- Vladimir Robles-Bykbaev
GI-IATa, UNESCO Chair on Support Technologies for Educational Inclusion, Universidad Politécnica Salesiana, Cuenca 010103, Ecuador
4 shared publications - Yaroslava Robles Bykbaev
Universidad Politécnica Salesiana
3 shared publications - Adolfo Jara-Gavilanes
Computer Science, Politecnica Salesiana University, Cuenca, Ecuador; Computer Science, SRH University Berlin, Berlin, Germany; Computer Science, Kadir Has University, Istanbul, Turkey, Ecuador
1 shared publication - Bryan Calle-Curillo
Universidad Politécnica Salesiana
1 shared publication - C. Japón-Chavez
Universidad Politécnica Salesiana
1 shared publication - D. Sánchez-Noboa
Universidad Politécnica Salesiana
1 shared publication
Latest Publications
- Open Access
- 16 September 2026
- Published by MDPI AG in Preprints.org
Mitigating LLM Prompt Injections via Multi-Vector Representation and Voting-Based Routing
Large Language Models (LLMs) suffer from the problem of prompt injection vulnerability, but typical one-step defenses generally fail at the task by having a high amount of false positives. We introduce an innovative two-sided defense architecture that applies semantic multi-vector diversification and ensemble modeling to provide LLM protection without the need for closed black-box solutions. In our framework, we use three different semantic spaces mapping via publicly available embedding algorithms, where each is then scored by parallel classifiers based on gradient boosting machines. The inputs go through an adaptive voting-based routing process; those that receive low entropy ratings are classified normally, while high entropy items are sent through a secondary classifier – a local LLM expert, the Llama Guard 3, that handles the input. Importantly, this solution uses a two-way protection method, which checks the generated outputs and catches late execution and context manipulation vectors. Our experiments on a 93K dataset show that the ensemble approach performed better than any individual sub-model, reaching 90.69 % accuracy and 94.11 % specificity, minimizing any user experience disturbances. The multi-vector layer demonstrated high empirical resilience to tokenization errors, while the output layer filtered out many bypassed malicious items.
- 16 September 2026
- Published by MDPI AG in Preprints.org
Open AccessMitigating LLM Prompt Injections via Multi-Vector Representation and Voting-Based Routing
Large Language Models (LLMs) suffer from the problem of prompt injection vulnerability, but typical one-step defenses generally fail at the task by having a high amount of false positives. We introduce an innovative two-sided defense architecture that applies semantic multi-vector diversification and ensemble modeling to provide LLM protection without the need for closed black-box solutions. In our framework, we use three different semantic spaces mapping via publicly available embedding algorithms, where each is then scored by parallel classifiers based on gradient boosting machines. The inputs go through an adaptive voting-based routing process; those that receive low entropy ratings are classified normally, while high entropy items are sent through a secondary classifier – a local LLM expert, the Llama Guard 3, that handles the input. Importantly, this solution uses a two-way protection method, which checks the generated outputs and catches late execution and context manipulation vectors. Our experiments on a 93K dataset show that the ensemble approach performed better than any individual sub-model, reaching 90.69 % accuracy and 94.11 % specificity, minimizing any user experience disturbances. The multi-vector layer demonstrated high empirical resilience to tokenization errors, while the output layer filtered out many bypassed malicious items.
- 27 April 2025
- Published by Springer Nature
- p. 150-159
Design of a Prototype for Measuring Electrodermal Signals Through an Ancestral Medicine Approach
- 27 April 2025
- Published by Springer Nature
- p. 150-159
Design of a Prototype for Measuring Electrodermal Signals Through an Ancestral Medicine Approach
- 1 August 2018
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1-4
An educational support tool based on robotic assistants, mobile apps, and expert systems for children with Down syndrome
The World Health Organization (WHO) claims that each year approximately 303,000 newborns die due congenital anomalies. These such of diseases can produce long-term disabilities, and the three most critical congenital anomalies are heart defects, neural tube defects, and Down syndrome (DS). Ecuador is a country with a high prevalence of DS, since 1 out of every 700 children born with this disease. On the other hand, worldwide, the number of children born with DS is 1 per 1,000. For these reasons, in this paper, we present a technological tool aimed at supporting the school curriculum for children from 6 to 12 years with DS. This tool relies on a low-cost robotic assistant, a mobile app, and a module to suggest educational activities for the children. With the aim of determining the children's response to the robot as well as to the mobile application, we have conducted a first pilot experiment with 21 children (without disabilities). Each of the participating children interacted with the system, and after this activity, they responded a survey. The initial results have shown that children are very interested and motivated to perform educational activities with the robot.
- 1 August 2018
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1-4
An educational support tool based on robotic assistants, mobile apps, and expert systems for children with Down syndrome
The World Health Organization (WHO) claims that each year approximately 303,000 newborns die due congenital anomalies. These such of diseases can produce long-term disabilities, and the three most critical congenital anomalies are heart defects, neural tube defects, and Down syndrome (DS). Ecuador is a country with a high prevalence of DS, since 1 out of every 700 children born with this disease. On the other hand, worldwide, the number of children born with DS is 1 per 1,000. For these reasons, in this paper, we present a technological tool aimed at supporting the school curriculum for children from 6 to 12 years with DS. This tool relies on a low-cost robotic assistant, a mobile app, and a module to suggest educational activities for the children. With the aim of determining the children's response to the robot as well as to the mobile application, we have conducted a first pilot experiment with 21 children (without disabilities). Each of the participating children interacted with the system, and after this activity, they responded a survey. The initial results have shown that children are very interested and motivated to perform educational activities with the robot.