collaborators

13 papers

cs.LG2026

Don't Go Breaking My LLM: The Impact of Pruning Attention Layers on Explanation Faithfulness and Confidence Calibration

Pietro Tropeano, Maria Maistro, Tuukka Ruotsalo +1

Pruning Large Language Models (LLMs) reduces memory and inference costs by removing parts of the network, producing smaller models that retain most of their accuracy. As attention…

cs.CL2026

How Context Shapes Truth: Geometric Transformations of Statement-level Truth Representations in LLMs

Shivam Adarsh, Maria Maistro, Christina Lioma

Large Language Models (LLMs) often encode whether a statement is true as a vector in their residual stream activations. These vectors, also known as truth vectors, have been studie…

cs.CL2026

Correcting Gradient-Based Circuit Localization via Interaction-Aware Backpropagation

Joakim Edin, Casper L. Christensen, Róbert Csordás +5

Circuit localization methods aim to identify the subset of model components responsible for specific behaviors in large language models, enabling detailed mechanistic analysis. Mos…

cs.IR2026

Can Fairness Be Prompted? Prompt-Based Debiasing Strategies in High-Stakes Recommendations

Mihaela Rotar, Theresia Veronika Rampisela, Maria Maistro

Large Language Models (LLMs) can infer sensitive attributes such as gender or age from indirect cues like names and pronouns, potentially biasing recommendations. While several deb…

cs.IR2026

Post-Training Denoising of User Profiles with LLMs in Collaborative Filtering Recommendation

Ervin Dervishaj, Maria Maistro, Tuukka Ruotsalo +1

Implicit feedback -- the main data source for training Recommender Systems (RSs) -- is inherently noisy and has been shown to negatively affect recommendation effectiveness. Denois…

cs.CY2026

Measuring Individual User Fairness with User Similarity and Effectiveness Disparity

Theresia Veronika Rampisela, Maria Maistro, Tuukka Ruotsalo +1

Individual user fairness is commonly understood as treating similar users similarly. In Recommender Systems (RSs), several evaluation measures exist for quantifying individual user…