Massimo Rizzoli
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Massimo Rizzoli has a total of 7 co-authors. Here are the top co-authors by number of shared publications:
- Giuseppe Riccardi
Signals and Interactive Systems Lab University of Trento Italy
4 shared publications - Simone Alghisi
Signals and Interactive Systems Lab University of Trento Italy
3 shared publications Gabriel Roccabruna
Amazon
2 shared publications- Gabriel Roccabruna
Signals and Interactive Systems Lab University of Trento Italy
2 shared publications Seyed Mahed Mousavi
Signals and Interactive Systems Lab University of Trento Italy
2 shared publicationsOlha Khomyn
Signals and Interactive Systems Lab University of Trento Italy
1 shared publication
Latest Publications
- Open Access
[De|Re]constructing VLMs' Reasoning in Counting
Vision-Language Models (VLMs) have recently gained attention due to their competitive performance on multiple downstream tasks, achieved by following user-input instructions. However, VLMs still exhibit several limitations in visual reasoning, such as difficulties in identifying relations (e.g., spatial, temporal, and among objects), understanding temporal sequences (e.g., frames), and counting objects. In this work, we go beyond score-level benchmark evaluations of VLMs by investigating the underlying causes of their failures and proposing a targeted approach to improve their reasoning capabilities. We study the reasoning skills of seven state-of-the-art VLMs in the counting task under controlled experimental conditions. Our experiments show that VLMs are highly sensitive to the number and type of objects, their spatial arrangement, and the co-occurrence of distractors. A layer-wise analysis reveals that errors are due to incorrect mapping of the last-layer representation into the output space. Our targeted training shows that fine-tuning just the output layer improves accuracy by up to 21%. We corroborate these findings by achieving consistent improvements on real-world datasets.
Open Access[De|Re]constructing VLMs' Reasoning in Counting
Vision-Language Models (VLMs) have recently gained attention due to their competitive performance on multiple downstream tasks, achieved by following user-input instructions. However, VLMs still exhibit several limitations in visual reasoning, such as difficulties in identifying relations (e.g., spatial, temporal, and among objects), understanding temporal sequences (e.g., frames), and counting objects. In this work, we go beyond score-level benchmark evaluations of VLMs by investigating the underlying causes of their failures and proposing a targeted approach to improve their reasoning capabilities. We study the reasoning skills of seven state-of-the-art VLMs in the counting task under controlled experimental conditions. Our experiments show that VLMs are highly sensitive to the number and type of objects, their spatial arrangement, and the co-occurrence of distractors. A layer-wise analysis reveals that errors are due to incorrect mapping of the last-layer representation into the output space. Our targeted training shows that fine-tuning just the output layer improves accuracy by up to 21%. We corroborate these findings by achieving consistent improvements on real-world datasets.
- Open Access
- 1 January 2025
- Published by Association for Computational Linguistics (ACL)
- p. 4462-4480
CIVET: Systematic Evaluation of Understanding in VLMs
Massimo Rizzoli, Simone Alghisi, Olha Khomyn, Gabriel Roccabruna, Seyed Mahed Mousavi, Giuseppe Riccardi. Findings of the Association for Computational Linguistics: EMNLP 2025. 2025.
- 1 January 2025
- Published by Association for Computational Linguistics (ACL)
- p. 4462-4480
Open AccessCIVET: Systematic Evaluation of Understanding in VLMs
Massimo Rizzoli, Simone Alghisi, Olha Khomyn, Gabriel Roccabruna, Seyed Mahed Mousavi, Giuseppe Riccardi. Findings of the Association for Computational Linguistics: EMNLP 2025. 2025.
- Open Access
- 1 January 2024
- Published by Association for Computational Linguistics (ACL)
- p. 180-197
Should We Fine-Tune or RAG? Evaluating Different Techniques to Adapt LLMs for Dialogue
Simone Alghisi, Massimo Rizzoli, Gabriel Roccabruna, Seyed Mahed Mousavi, Giuseppe Riccardi. Proceedings of the 17th International Natural Language Generation Conference. 2024.
- 1 January 2024
- Published by Association for Computational Linguistics (ACL)
- p. 180-197
Open AccessShould We Fine-Tune or RAG? Evaluating Different Techniques to Adapt LLMs for Dialogue
Simone Alghisi, Massimo Rizzoli, Gabriel Roccabruna, Seyed Mahed Mousavi, Giuseppe Riccardi. Proceedings of the 17th International Natural Language Generation Conference. 2024.
- Open Access
- 1 January 2024
- Published by Association for Computational Linguistics (ACL)
- p. 20402-20415
Will LLMs Replace the Encoder-Only Models in Temporal Relation Classification?
Gabriel Roccabruna, Massimo Rizzoli, Giuseppe Riccardi. Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing. 2024.
- 1 January 2024
- Published by Association for Computational Linguistics (ACL)
- p. 20402-20415
Open AccessWill LLMs Replace the Encoder-Only Models in Temporal Relation Classification?
Gabriel Roccabruna, Massimo Rizzoli, Giuseppe Riccardi. Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing. 2024.