Research Work

Papers

• Francesco Calcagno, Luca Serfilippi, Giorgio Franceschelli, Marco Garavelli, Mirco Musolesi and Ivan Rivalta. 2026. Quantum Chemistry-Driven Molecular Inverse Design of Stable Isomers with Data-Free Reinforcement Learning. Journal of Chemical Theory and Computation. doi.org/10.1021/acs.jctc.5c02055
  The inverse design (ID) of molecules remains one of the greatest challenges in chemistry. Machine learning and artificial intelligence (AI) methods are increasingly employed to generate candidate molecules with tailored properties but mostly rely on pretraining over large data sets, which introduces bias. Here, we present a data-free generative AI model called PROTEUS that integrates reinforcement learning with on-the-fly quantum mechanical calculations to enable the de novo design of molecules from first-principles. The AI tool uses a custom syntax and hierarchical learning architecture to navigate the chemical space without prior knowledge, optimizing the desired chemical property. We demonstrate the efficiency of our software by solving complex molecular design tasks related to the maximization of isomerization energy gaps for styrene derivatives. By solving ID problems for which the exact solutions are known, PROTEUS proved to be robust and flexible enough to perform a broad exploration of different chemical spaces while successfully exploiting chemical rewards. This framework opens new avenues for quantum chemistry-driven unbiased molecular design, offering a flexible and scalable strategy to address design challenges in chemistry.
• Luca Serfilippi, Giorgio Franceschelli, Antonio Corradi and Mirco Musolesi. 2025. Complexity-Regularized Proximal Policy Optimization. arXiv:2509.20509 [cs.LG]
  Policy gradient methods usually rely on entropy regularization to prevent premature convergence. However, maximizing entropy indiscriminately pushes the policy towards a uniform distribution, often overriding the reward signal if not optimally tuned. We propose replacing the standard entropy term with a self-regulating complexity term, defined as the product of Shannon entropy and disequilibrium, where the latter quantifies the distance from the uniform distribution. Unlike pure entropy, which favors maximal disorder, this complexity measure is zero for both fully deterministic and perfectly uniform distributions, i.e., it is strictly positive for systems that exhibit a meaningful interplay between order and randomness. These properties ensure the policy maintains beneficial stochasticity while reducing regularization pressure when the policy is highly uncertain, allowing learning to focus on reward optimization. We introduce Complexity-Regularized Proximal Policy Optimization (CR-PPO), a modification of PPO that leverages this dynamic. We empirically demonstrate that CR-PPO is significantly more robust to hyperparameter selection than entropy-regularized PPO, achieving consistent performance across orders of magnitude of regularization coefficients and remaining harmless when regularization is unnecessary, thereby reducing the need for expensive hyperparameter tuning.
• Giorgio Franceschelli and Mirco Musolesi. 2025. DiffSampling: Enhancing Diversity and Accuracy in Neural Text Generation. Transactions on Machine Learning Research. PDF Code Video
  Despite their growing capabilities, language models still frequently reproduce content from their training data, generate repetitive text, and favor common grammatical patterns and vocabulary. A possible cause is the decoding strategy: the most common strategies either consider only the most probable tokens, which reduces output diversity, or increase the likelihood of unlikely tokens, compromising output accuracy and correctness. In this paper, we propose DiffSampling, a new decoding method that leverages a mathematical analysis of the token probability distribution to ensure the generation of contextually appropriate text. In particular, the difference between consecutive, sorted probabilities can be used to truncate incorrect tokens. In addition, we also propose two variations of the proposed method that aim to correct the subtle inconsistencies of common sampling strategies. Experiments involving four different text-generation tasks demonstrate that our approach consistently performs at least on par with the existing methods it builds upon in terms of quality, despite sampling from a larger set of tokens.
• Giorgio Franceschelli and Mirco Musolesi. 2025. Thinking Outside the (Gray) Box: A Context-Based Score for Assessing Value and Originality in Neural Text Generation. arXiv:2502.13207 [cs.CL]
  Despite the increasing use of large language models for creative tasks, their outputs often lack diversity. Common solutions, such as sampling at higher temperatures, can compromise the quality of the results. Drawing on information theory, we propose a context-based score to quantitatively evaluate value and originality. This score incentivizes accuracy and adherence to the request while fostering divergence from the learned distribution. We propose using our score as a reward in a reinforcement learning framework to fine-tune large language models for maximum performance. We validate our strategy through experiments in poetry generation and math problem solving, demonstrating that it enhances the value and originality of the generated solutions.
• Giorgio Franceschelli, Claudia Cevenini and Mirco Musolesi. 2024. Training Foundation Models as Data Compression: On Information, Model Weights and Copyright Law. arXiv:2407.13493 [cs.CY]
  The training process of foundation models as for other classes of deep learning systems is based on minimizing the reconstruction error over a training set. For this reason, they are susceptible to the memorization and subsequent reproduction of training samples. In this paper, we introduce a training-as-compressing perspective, wherein the model’s weights embody a compressed representation of the training data. From a copyright standpoint, this point of view implies that the weights could be considered a reproduction or a derivative work of a potentially protected set of works. We investigate the technical and legal challenges that emerge from this framing of the copyright of outputs generated by foundation models, including their implications for practitioners and researchers. We demonstrate that adopting an information-centric approach to the problem presents a promising pathway for tackling these emerging complex legal issues.
• Giorgio Franceschelli and Mirco Musolesi. 2024. Creative Beam Search: LLM-as-a-Judge For Improving Response Generation. Proc. of the 15th International Conference on Computational Creativity (ICCC’24). Jonkoping, Sweden. PDF
  Large language models are revolutionizing several areas, including artificial creativity. However, the process of generation in machines profoundly diverges from that observed in humans. In particular, machine generation is characterized by a lack of intentionality and an underlying creative process. We propose a method called Creative Beam Search that uses Diverse Beam Search and LLM-as-a-Judge to perform response generation and response validation. The results of a qualitative experiment show how our approach can provide better output than standard sampling techniques. We also show that the response validation step is a necessary complement to the response generation step.
• Giorgio Franceschelli and Mirco Musolesi. 2024. Do Agents Dream of Electric Sheep?: Improving Generalization in Reinforcement Learning through Generative Learning. arXiv:2403.07979 [cs.LG]
  The Overfitted Brain hypothesis suggests dreams happen to allow generalization in the human brain. Here, we ask if the same is true for reinforcement learning agents as well. Given limited experience in a real environment, we use imagination-based reinforcement learning to train a policy on dream-like episodes, where non-imaginative, predicted trajectories are modified through generative augmentations. Experiments on four ProcGen environments show that, compared to classic imagination and offline training on collected experience, our method can reach a higher level of generalization when dealing with sparsely rewarded environments.
• Giorgio Franceschelli and Mirco Musolesi. 2024. Reinforcement Learning for Generative AI: State of the Art, Opportunities and Open Research Challenges. Journal of Artificial Intelligence Research, 79, 417-446. doi.org/10.1613/jair.1.15278
  Generative Artificial Intelligence (AI) is one of the most exciting developments in Computer Science of the last decade. At the same time, Reinforcement Learning (RL) has emerged as a very successful paradigm for a variety of machine learning tasks. In this survey, we discuss the state of the art, opportunities and open research questions in applying RL to generative AI. In particular, we will discuss three types of applications, namely, RL as an alternative way for generation without specified objectives; as a way for generating outputs while concurrently maximizing an objective function; and, finally, as a way of embedding desired characteristics, which cannot be easily captured by means of an objective function, into the generative process. We conclude the survey with an in-depth discussion of the opportunities and challenges in this fascinating emerging area.
• Giorgio Franceschelli and Mirco Musolesi. 2023. On the Creativity of Large Language Models. AI & SOCIETY, 40, 3785-3795. doi.org/10.1007/s00146-024-02127-3
  Large Language Models (LLMs) are revolutionizing several areas of Artificial Intelligence. One of the most remarkable applications is creative writing, e.g., poetry or storytelling: the generated outputs are often of astonishing quality. However, a natural question arises: can LLMs be really considered creative? In this article we firstly analyze the development of LLMs under the lens of creativity theories, investigating the key open questions and challenges. In particular, we focus our discussion around the dimensions of value, novelty and surprise as proposed by Margaret Boden in her work. Then, we consider different classic perspectives, namely product, process, press and person. We discuss a set of easy'' andhard’’ problems in machine creativity, presenting them in relation to LLMs. Finally, we examine the societal impact of these technologies with a particular focus on the creative industries, analyzing the opportunities offered by them, the challenges arising by them and the potential associated risks, from both legal and ethical points of view.
• Giorgio Franceschelli and Mirco Musolesi. 2022. DeepCreativity: Measuring Creativity with Deep Learning Techniques. Intelligenza Artificiale, 16(2), 151-163. doi:10.3233/IA-220136
  Measuring machine creativity is one of the most fascinating challenges in Artificial Intelligence. This paper explores the possibility of using generative learning techniques for automatic assessment of creativity. The proposed solution does not involve human judgement, it is modular and of general applicability. We introduce a new measure, namely DeepCreativity, based on Margaret Boden’s definition of creativity as composed by value, novelty and surprise. We evaluate our methodology (and related measure) considering a case study, i.e., the generation of 19th century American poetry, showing its effectiveness and expressiveness.
• Giorgio Franceschelli and Mirco Musolesi. 2022. Copyright in Generative Deep Learning. Data & Policy, 4, E17. doi:10.1017/dap.2022.10
  Machine-generated artworks are now part of the contemporary art scene: they are attracting significant investments and they are presented in exhibitions together with those created by human artists. These artworks are mainly based on generative deep learning (GDL) techniques, which have seen a formidable development and remarkable refinement in the very recent years. Given the inherent characteristics of these techniques, a series of novel legal problems arise. In this article, we consider a set of key questions in the area of GDL for the arts, including the following: is it possible to use copyrighted works as training set for generative models? How do we legally store their copies in order to perform the training process? Who (if someone) will own the copyright on the generated data? We try to answer these questions considering the law in force in both the United States and the European Union, and potential future alternatives. We then extend our analysis to code generation, which is an emerging area of GDL. Finally, we also formulate a set of practical guidelines for artists and developers working on deep learning generated art, as well as some policy suggestions for policymakers.
• Giorgio Franceschelli and Mirco Musolesi. 2024. Creativity and Machine Learning: A Survey. ACM Computing Surveys, 56(11), 283:1-41. doi:10.1145/3664595
  There is a growing interest in the area of machine learning and creativity. This survey presents an overview of the history and the state of the art of computational creativity theories, key machine learning techniques (including generative deep learning), and corresponding automatic evaluation methods. After presenting a critical discussion of the key contributions in this area, we outline the current research challenges and emerging opportunities in this field.

Books and Contributions

• Giorgio Franceschelli, I, Artist. Opere d’arte e intelligenza artificiale: il curioso caso del diritto d’autore, Ventura Edizioni, Senigallia, 2019.
• Giorgio Franceschelli, L’importanza del suono nel progresso culturale, in Prospettive sonore. Percezione e mediazione, by A. Calanchi e A. Laquidara, Galaad, Giulianova, 2016, pp. 257 - 274.