Students

Open projects for master theses

If you are interested in working with me for your master theses send me an email at umili@diag.uniroma1.it

You can either propose your own idea for the theses or pick one project from the list below.

• Logically informed autoregressive learning. Training/ finetuning autoregresive deep models (LLM / RL agents ) with background knowledge in formal languages. (References: paper4.pdf, 2504.13139, 2311.00094, 2312.03905)

• RL for logically constrained sequence generation. Training sequence generators with reinforcement learning using synthetic rewards coming from logical temporal specifications (References: 2106.01345, trajectory-transformer)

• Natural language instructions to policies via LLMs+NeSy RL. Training RL agents capable of following natural language instructions by first turning them into formal specifications that are then solved via NeSy RL. (References: 2010.03950, 1807.06333, 2102.06858)

Past theses