Internships

The IDA Unit of the Digital Health and Well Being Center is looking for a bachelor/master student interested in pursuing an internship in knowledge engineering, with a specific focus on multi-modal knowledge graphs (MMKG) for healthcare. The internship will be centered on work related to the Multi-Modal Knowledge Graph Supporting Personalized Health (FuS-KG), currently under development within the IDA group.

The ideal candidate is a proactive and dynamic young person with good organizational and interpersonal skills, with a propensity for teamwork, motivated to undertake a training experience in an international context.

The intern, working alongside the Research Unit staff, will mainly contribute to the development of the following activities:  

• Revising the existing content of FuS-KG, such as integrating recipes with nutritional information.

• Extending the FuS-KG resource with new data related to health related domains, including (but not limited to) food, recipes, daily and physical activities.

• Aligning FuS-KG with state-of-the-art ontologies by studying ontology matching tools and performing both automatic and manual matching.

• Improving the internal Python pipeline for the materialization and population of the MMKG

• Analyzing state-of-the-art deep learning models for feature extraction and incorporating features into FuS-KG nodes.

MORE INFO 👉 https://jobs.fbk.eu/Annunci/Offerte_di_lavoro_MULTI_MODAL_KNOWLEDGE_GRAPH_FOR_HEALTHCARE_239305325.htm

Planned activities

The Research Unit is looking for a Master’s student interested in pursuing an internship experience on Agent Aggregation Strategies in Multi-Agent Large Language Model Systems.

Multi-agent systems based on Large Language Models often rely on multiple agents that analyze the same problem from different perspectives. However, the final quality of the system does not depend only on the quality of the individual agents, but also on how their outputs are combined, compared, weighted, revised, or selected.

For this reason, aggregation is a crucial bottleneck in multi-agent reasoning architectures. A weak aggregation strategy can discard correct expert outputs, amplify systematic errors, or introduce unnecessary variability, even when some agents have produced useful reasoning. Conversely, a robust aggregation mechanism can improve accuracy, support error correction, and make the overall decision process more interpretable.

The internship will focus on the Pool of Experts (PoE) framework, a multi-agent architecture in which several expert agents contribute to a shared reasoning process. The main goal is to investigate how different aggregation methods influence final system performance and whether more sophisticated aggregation strategies can improve accuracy, robustness, explainability, and consistency across tasks.

The work will compare simple aggregation methods, such as majority voting, with more advanced approaches based on deliberation, confidence estimation, role-aware weighting, disagreement analysis, meta-reasoning, and final decision-making agents. Particular attention will be given to understanding when aggregation improves the system, when it introduces new errors, and how it can be made more reliable.

Main activities include:

• Conducting a literature review on aggregation methods in multi-agent systems and Large Language Model reasoning;

• Analyzing how individual expert agents contribute to the final decision;

• Comparing aggregation methods such as majority voting, weighted voting, ranking-based selection, confidence-based aggregation, and deliberative final decision-making;

• Investigating disagreement among agents as a signal for uncertainty, ambiguity, or task difficulty;

• Designing and implementing new aggregation strategies within the PoE system;

• Evaluating whether aggregation methods recover individual agent errors or introduce new mistakes;

• Measuring the impact of aggregation on accuracy, robustness, interpretability, and consistency across tasks.

This internship offers experience at the frontier of Large Language Models, multi-agent reasoning, decision aggregation, and trustworthy AI evaluation.

Requirements:

• Enrollment in a Master’s degree in one of the following fields: Computer Science, Data Science, Artificial Intelligence, Cognitive Sciences, or related disciplines;

• Fluent knowledge of English, minimum level B2;

• Strong interpersonal and teamwork skills;

• Good adaptability, initiative, and autonomy;

• Proficiency in Python;

• Familiarity with Large Language Models, generative AI, and model evaluation;

• Interest in multi-agent systems, reasoning, decision-making, or evaluation methodologies.

Additional knowledge of open-weight LLMs, prompt engineering, and statistical evaluation will be considered a plus.

Internship information:

• Internship start date: second semester 2026;

• Internship experience duration: 6 months;

• Opportunity reserved for Master’s students;

• The internship experience will take place at FBK;

• We offer support in finding accommodation at affiliated facilities, in the case of off-site candidates;

• We offer the possibility to use the internal canteen service;

• Possible recognition of participation allowance.

How to apply:

Interested candidates are asked to send their application by email to:

Patrizio Bellan - pbellan@fbk.eu

The application must include the following documents in .pdf format:

• curriculum vitae;

• motivation letter.

The motivation letter is required to apply.

For any further information, please contact the Human Resources Department at: jobs@fbk.eu

Application deadline: the notice will be withdrawn when the desired applications are reached.

Planned activities

The Research Unit is looking for a Master’s student interested in pursuing an internship experience on Self-Improving Large Language Models through Multi-Agent Reasoning.

The internship will explore how a general-purpose Large Language Model, such as Gemma or similar open-weight models, can improve its reasoning performance without relying on external training data. The work will build on the Pool of Experts (PoE) framework, a multi-agent architecture in which several expert agents analyze a problem from different perspectives and cooperate to produce a final decision.

The main goal of the internship is to investigate whether PoE can be used not only as an inference-time reasoning framework, but also as a mechanism for model self-improvement. Starting from a generalist LLM, the intern will study how the multi-agent system can be used to improve model performance without introducing new annotated datasets.

Main activities include:

• Exploring self-improvement strategies based on internal model outputs, such as self-consistency, self-reflection, critique, debate, and self-distillation;

• Designing experiments in which PoE agents generate, evaluate, revise, or select improved reasoning paths;

• Investigating whether synthetic traces produced by the model itself can support lightweight adaptation or improved inference strategies;

• Comparing baseline single-model performance with PoE-based self-enhancement strategies.
  
  

This internship offers experience at the frontier of Large Language Models, multi-agent systems, self-improving AI, and reasoning-oriented model evaluation

Requirements:

• Enrollment in a Master’s degree in one of the following fields: Computer Science, Data Science, Artificial Intelligence, Cognitive Sciences, or related disciplines;

• Fluent knowledge of English, minimum level B2;

• Strong interpersonal and teamwork skills;

• Good adaptability, initiative, and autonomy;

• Proficiency in Python;

• Familiarity with Large Language Models, generative AI, and model training;

• Interest in multi-agent systems, reasoning, model evaluation, or efficient adaptation techniques.

Additional knowledge of open-weight LLMs, LoRA, prompt engineering, or evaluation benchmarks will be considered a plus.

Internship information:

• Internship start date: second semester 2026;

• Internship experience duration: 6 months;

• Opportunity reserved for Master’s students;

• The internship experience will take place at FBK;

• We offer support in finding accommodation at affiliated facilities, in the case of off-site candidates;

• We offer the possibility to use the internal canteen service;

• Possible recognition of participation allowance.

How to apply:

Interested candidates are asked to send their application by email to:

Patrizio Bellan - pbellan@fbk.eu

The application must include the following documents in .pdf format:

• curriculum vitae;

• motivation letter.

The motivation letter is required to apply.

For any further information, please contact the Human Resources Department at: jobs@fbk.eu

Application deadline: the notice will be withdrawn when the desired applications are reached.

Planned activities

The Research Unit is looking for a Master’s student interested in pursuing an internship experience on Model Forgetting and Knowledge Removal in Large Language Models.

Large Language Models can store or reproduce information acquired during training, including outdated facts, incorrect knowledge, biased associations, or personal and sensitive information. This raises an important research question: how can a model be made to forget specific information while preserving its general capabilities?

The internship will investigate approaches for removing, suppressing, or editing specific knowledge in LLMs. Starting from an open-weight general-purpose model, such as Gemma or similar architectures, the intern will explore methods to identify information that should no longer be produced by the model and evaluate strategies to reduce or eliminate its availability in model outputs.

The work will focus on both technical and practical aspects of model forgetting. Relevant applications include the correction of memorized false information, the removal of personal or sensitive data, compliance with privacy requirements, and the development of safer and more controllable AI systems.

Main activities include:

• Conducting a literature review on model forgetting, machine unlearning, knowledge editing, and privacy-preserving LLM adaptation;

• Designing experiments to test whether a model has memorized or can reproduce target information;

• Exploring methods for knowledge removal, such as fine-tuning-based unlearning, negative preference optimization, knowledge editing, or prompt-level suppression;

• Evaluating the trade-off between forgetting target information and preserving the model’s general performance;

• Implementing experiments on open-weight LLMs, such as Gemma or similar models;

• Defining evaluation protocols to measure forgetting, residual memorization, and unintended degradation of model capabilities.

This internship offers experience at the frontier of Large Language Models, trustworthy AI, privacy-aware machine learning, and controllable model adaptation.

Requirements:

• Enrollment in a Master’s degree in one of the following fields: Computer Science, Data Science, Artificial Intelligence, Cognitive Sciences, or related disciplines;

• Fluent knowledge of English, minimum level B2;

• Strong interpersonal and teamwork skills;

• Good adaptability, initiative, and autonomy;

• Proficiency in Python;

• Familiarity with Large Language Models, generative AI, and model training;

• Interest in trustworthy AI, privacy, model evaluation, or efficient adaptation techniques.

Additional knowledge of open-weight LLMs, Hugging Face, PyTorch, LoRA, model fine-tuning, knowledge editing, or machine unlearning will be considered a plus.

Internship information:

• Internship start date: second semester 2026;

• Internship experience duration: 6 months;

• Opportunity reserved for Master’s students;

• The internship experience will take place at FBK;

• We offer support in finding accommodation at affiliated facilities, in the case of off-site candidates;

• We offer the possibility to use the internal canteen service;

• Possible recognition of participation allowance.

How to apply:

Interested candidates are asked to send their application by email to:

Patrizio Bellan - pbellan@fbk.eu

The application must include the following documents in .pdf format:

• curriculum vitae;

• motivation letter.

The motivation letter is required to apply.

For any further information, please contact the Human Resources Department at: jobs@fbk.eu

Application deadline: the notice will be withdrawn when the desired applications are reached.