Rodrigo Alvarez
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Rodrigo Alvarez has a total of 16 co-authors. Here are the top co-authors by number of shared publications:
- Mina Soroberto
Research Group, University of Salamanca (PIC 999846610),Edificio Multiusos I+D+i, Calle Espejo s/n, Salamanca, 37007, Spain
3 shared publications - Ana Delgado-García
Universidad de Salamanca
2 shared publications - Fernando De la Prieta
University of Salamanca
2 shared publications - Javier Prieto
BISITE Research Group, University of Salamanca, Salamanca, Salamanca, 37007, Spain
2 shared publications Mario Cartón
MADISON MK, Edificio Madison Arena, C/ Enrique Cubero 9, Valladolid, 47014, Spain
2 shared publications- Abir Rebei
University of Salamanca
1 shared publication
Latest Publications
- 17 October 2026
- Published by Springer Nature
- p. 3-14
A Lightweight Dispersion Metric Suite for Real-Time Convergence Monitoring in Synthetic Populations
- 17 October 2026
- Published by Springer Nature
- p. 3-14
A Lightweight Dispersion Metric Suite for Real-Time Convergence Monitoring in Synthetic Populations
- Open Access
- 8 August 2026
- Published by CERN in Open Research Europe
A lightweight computational method for monitoring lexical response convergence in LLM-based synthetic populations
Background LLM-based synthetic populations are increasingly used to simulate open-ended survey, interview, user-research, and group-style responses. These workflows can support early exploratory research, but they also create a methodological risk: agents configured as different participants may nevertheless converge towards similar wording, examples, or answer templates. Without an explicit monitoring step, apparent agreement can be difficult to distinguish from model-level homogenization. Methods We present a lightweight computational method for monitoring lexical response convergence in generated agent populations. The method uses a platform-independent response-record schema, deterministic normalization, pairwise lexical overlap, vocabulary-diversity diagnostics, data-quality counters, threshold-based inspection flags, and reproducible JSON reporting. The reference implementation is the Python package synthetic-response-metrics version 0.2.5, which supports CSV, JSON, JSONL, API, and command-line execution. Worked examples and initial validation We provide initial validation using controlled synthetic examples created for this article. The validation includes a data-quality example, low-, moderate-, and high-convergence scenarios, a population-size sensitivity analysis, and a high-convergence diagnostic pair. These examples show expected metric behaviour, including increasing overlap under controlled convergence, changing pair counts as population size increases, and traceability of alerts through top similar pairs. In addition, an applied example generated from a synthetic-population platform uses 10 synthetic agents and three open-ended questions to check operational portability of the same input schema. Conclusions The method provides a reproducible lexical baseline for observing whether generated agents are producing overly similar answers. It does not replace semantic evaluation, human baselines, or domain validation, and the default threshold is provisional. However, it offers a practical monitoring layer that can be embedded in synthetic-population studies before more expensive or domain-specific validation is available.
- 8 August 2026
- Published by CERN in Open Research Europe
Open AccessA lightweight computational method for monitoring lexical response convergence in LLM-based synthetic populations
Background LLM-based synthetic populations are increasingly used to simulate open-ended survey, interview, user-research, and group-style responses. These workflows can support early exploratory research, but they also create a methodological risk: agents configured as different participants may nevertheless converge towards similar wording, examples, or answer templates. Without an explicit monitoring step, apparent agreement can be difficult to distinguish from model-level homogenization. Methods We present a lightweight computational method for monitoring lexical response convergence in generated agent populations. The method uses a platform-independent response-record schema, deterministic normalization, pairwise lexical overlap, vocabulary-diversity diagnostics, data-quality counters, threshold-based inspection flags, and reproducible JSON reporting. The reference implementation is the Python package synthetic-response-metrics version 0.2.5, which supports CSV, JSON, JSONL, API, and command-line execution. Worked examples and initial validation We provide initial validation using controlled synthetic examples created for this article. The validation includes a data-quality example, low-, moderate-, and high-convergence scenarios, a population-size sensitivity analysis, and a high-convergence diagnostic pair. These examples show expected metric behaviour, including increasing overlap under controlled convergence, changing pair counts as population size increases, and traceability of alerts through top similar pairs. In addition, an applied example generated from a synthetic-population platform uses 10 synthetic agents and three open-ended questions to check operational portability of the same input schema. Conclusions The method provides a reproducible lexical baseline for observing whether generated agents are producing overly similar answers. It does not replace semantic evaluation, human baselines, or domain validation, and the default threshold is provisional. However, it offers a practical monitoring layer that can be embedded in synthetic-population studies before more expensive or domain-specific validation is available.
- 29 July 2026
- Published by Ediciones Universidad de Salamanca
- p. 109-114
PREDICCIÓN DEL RIESGO DE INCENDIOS FORESTALES MEDIANTE TELEDETECCIÓN E IA (2020-2025)
Se sintetizan los avances de 2020-2025 en predicción del riesgo de incendios forestales mediante teledetección y modelos de inteligencia artificial. Se resume el uso de índices espectrales (NDVI, EVI, NDMI), variables térmicas, topográficas y antrópicas, y se comparan líneas base (regresión, Random Forest, SVM) con arquitecturas profundas (CNN, ConvLSTM, Transformers). Se destacan resultados recientes con AUC entre 0.85 y 0.96 y F1 entre 0.67 y 0.85, así como la incorporación de autosupervisión y adaptación eficiente de parámetros para transferir modelos entre regiones. Se proponen buenas prácticas de evaluación (F1/AUC, RMSE/MAE y Kappa) y recomendaciones para despliegue operativo con datos multifuente y restricciones de costo computacional.
- 29 July 2026
- Published by Ediciones Universidad de Salamanca
- p. 109-114
PREDICCIÓN DEL RIESGO DE INCENDIOS FORESTALES MEDIANTE TELEDETECCIÓN E IA (2020-2025)
Se sintetizan los avances de 2020-2025 en predicción del riesgo de incendios forestales mediante teledetección y modelos de inteligencia artificial. Se resume el uso de índices espectrales (NDVI, EVI, NDMI), variables térmicas, topográficas y antrópicas, y se comparan líneas base (regresión, Random Forest, SVM) con arquitecturas profundas (CNN, ConvLSTM, Transformers). Se destacan resultados recientes con AUC entre 0.85 y 0.96 y F1 entre 0.67 y 0.85, así como la incorporación de autosupervisión y adaptación eficiente de parámetros para transferir modelos entre regiones. Se proponen buenas prácticas de evaluación (F1/AUC, RMSE/MAE y Kappa) y recomendaciones para despliegue operativo con datos multifuente y restricciones de costo computacional.
- Open Access
- 30 June 2026
- Published by CERN in Open Research Europe
- Vol. 6, 223
A lightweight computational method for monitoring response convergence in LLM-based synthetic populations
Background LLM-based synthetic populations are increasingly used to simulate open-ended survey, interview, user-research, and group-style responses. These workflows can support early exploratory research, but they also create a methodological risk: agents configured as different participants may nevertheless converge towards similar wording, examples, or answer templates. Without an explicit monitoring step, apparent agreement can be difficult to distinguish from model-level homogenization. Methods We present a lightweight computational method for monitoring lexical response convergence in generated agent populations. The method uses a platform-independent response-record schema, deterministic normalization, pairwise lexical overlap, vocabulary-diversity diagnostics, data-quality counters, threshold-based inspection flags, and reproducible JSON reporting. The reference implementation is the Python package synthetic-response-metrics version 0.2.5, which supports CSV, JSON, JSONL, API, and command-line execution. Worked examples and initial validation We provide initial validation using controlled synthetic examples created for this article. The validation includes a data-quality example, low-, moderate-, and high-convergence scenarios, a population-size sensitivity analysis, and a high-convergence diagnostic pair. These examples show expected metric behaviour, including increasing overlap under controlled convergence, changing pair counts as population size increases, and traceability of alerts through top similar pairs. In addition, an applied example generated from a synthetic-population platform uses 10 synthetic agents and three open-ended questions to check operational portability of the same input schema. Conclusions The method provides a reproducible lexical baseline for observing whether generated agents are producing overly similar answers. It does not replace semantic evaluation, human baselines, or domain validation, and the default threshold is provisional. However, it offers a practical monitoring layer that can be embedded in synthetic-population studies before more expensive or domain-specific validation is available.
- 30 June 2026
- Published by CERN in Open Research Europe
- Vol. 6, 223
Open AccessA lightweight computational method for monitoring response convergence in LLM-based synthetic populations
Background LLM-based synthetic populations are increasingly used to simulate open-ended survey, interview, user-research, and group-style responses. These workflows can support early exploratory research, but they also create a methodological risk: agents configured as different participants may nevertheless converge towards similar wording, examples, or answer templates. Without an explicit monitoring step, apparent agreement can be difficult to distinguish from model-level homogenization. Methods We present a lightweight computational method for monitoring lexical response convergence in generated agent populations. The method uses a platform-independent response-record schema, deterministic normalization, pairwise lexical overlap, vocabulary-diversity diagnostics, data-quality counters, threshold-based inspection flags, and reproducible JSON reporting. The reference implementation is the Python package synthetic-response-metrics version 0.2.5, which supports CSV, JSON, JSONL, API, and command-line execution. Worked examples and initial validation We provide initial validation using controlled synthetic examples created for this article. The validation includes a data-quality example, low-, moderate-, and high-convergence scenarios, a population-size sensitivity analysis, and a high-convergence diagnostic pair. These examples show expected metric behaviour, including increasing overlap under controlled convergence, changing pair counts as population size increases, and traceability of alerts through top similar pairs. In addition, an applied example generated from a synthetic-population platform uses 10 synthetic agents and three open-ended questions to check operational portability of the same input schema. Conclusions The method provides a reproducible lexical baseline for observing whether generated agents are producing overly similar answers. It does not replace semantic evaluation, human baselines, or domain validation, and the default threshold is provisional. However, it offers a practical monitoring layer that can be embedded in synthetic-population studies before more expensive or domain-specific validation is available.
- 2 January 2026
- Published by Springer Nature
- p. 269-278
Neural Embedding-Based Recommendation System for Carbon Credit Project Matchmaking
- 2 January 2026
- Published by Springer Nature
- p. 269-278
Neural Embedding-Based Recommendation System for Carbon Credit Project Matchmaking
- 2 January 2026
- Published by Springer Nature
- p. 245-255
Towards a Decentralized and Transparent Carbon Market: Blockchain and IoT in Carbon Farming
- 2 January 2026
- Published by Springer Nature
- p. 245-255
Towards a Decentralized and Transparent Carbon Market: Blockchain and IoT in Carbon Farming