Enrico Liscio

Enrico Liscio

Postdoctoral Researcher · TU Delft

CV (PDF)

Research

My research addresses how language models can align with human values, not as a static, one-off training objective, but as something that is context-dependent and shaped through interaction. During my PhD, I developed methods to identify value-relevant signals in text (value inference); as a postdoc, I studied how language models can generate outputs that reflect different value orientations (value expression). In parallel, I work on subjectivity in NLP, modeling variation in interpretation rather than reducing it to a single ground truth.

I was recently awarded a Veni, a prestigious Dutch personal grant for early-career researchers. My project, Inference-Time Value Alignment in Multi-Turn Language Model Interaction, develops methods to infer user values from noisy, evolving signals during a conversation; to represent them in a form a language model can use; and to express them stably in generated language throughout a multi-turn exchange.

Publications

2026

  • A. Arzberger, E. Liscio, M.L. Lupetti, I. Martinez de Rituerto de Troya, and J. Yang. 2026. Co-Constructing Alignment: A Participatory Approach to Situating AI Values. In FAccT. [paper]

  • U. Khurana, M. van der Meer, E. Liscio, A. Fokkens, P.K. Murukannaiah. 2026. Not All Subjectivity Is the Same! Defining Desiderata for the Evaluation of Subjectivity in NLP. EMNLP 2026 (Findings). [pre-print]

  • P. Blum, E. Liscio, R. Zhang, C. Figueroa, P.K. Murukannaiah. 2026. Reading Between the Signs: Predicting Future Suicidal Ideation from Adolescent Social Media Texts. NLP4PI @ EMNLP 2026. [pre-print]

  • R.A.C. Diaconescu, I. Slanina, A. Florea, A.B. Trache, M.E. Coroi, A. Arzberger, J. Yang, E. Liscio. 2026. Beyond a Global Norm: Personalizing Toxicity Sensitivity in Language Models Without Retraining. WOAH @ EMNLP 2026. [pre-print]

  • K.W. Guan, S. Goeschl, M. Amara, E. Liscio, M. Esherick, A. Kozlovski, J. Gonçalves, M. de Reuver, C. Figueroa. 2026. Operationalizing Governance for Youth-Facing LLM Well-being Support: A Community-Based Participatory Study. In FAccT. [paper]

2025

  • E. Liscio*, L.C. Siebert*, C.M. Jonker, P.K. Murukannaiah. 2025. Value Preferences Estimation and Disambiguation in Hybrid Participatory Systems. In Journal of Artificial Intelligence Research (JAIR). [paper] [slides]

  • E. Liscio, M. Lorandi, P.K. Murukannaiah. 2025. News is More than a Collection of Facts: Moral Frame Preserving News Summarization. In COLM 2025. [paper] [poster]

  • A. Homayounirad, E. Liscio, T. Wang, C.M. Jonker, L.C. Siebert. 2025. Will Annotators Disagree? Identifying Subjectivity in Value-Laden Arguments. In EMNLP 2025 (Findings). [paper]

  • A. Kuber, E. Liscio, R. Zhang, C. Figueroa, P.K. Murukannaiah. 2025. Signs of Struggle: Spotting Cognitive Distortions across Language and Register. In AACL 2025 (Findings). [paper]

  • A.N. Dobrinoiu, A.C. Marcu, A. Homayounirad, L.C. Siebert, E. Liscio. 2025. Taking a SEAT: Predicting Value Interpretations from Sentiment, Emotion, Argument, and Topic Annotations. In VALE @ ECAI 2025. [paper] [slides] [poster]

2024

  • J. Park*, E. Liscio*, P.K. Murukannaiah. 2024. Morality is Non-Binary: Building a Pluralist Moral Sentence Embedding Space using Contrastive Learning. In EACL 2024 (Findings). [paper] [poster] [slides] [extended abstract]

  • M. van der Meer, N. Falk, P.K. Murukannaiah, E. Liscio. 2024. Annotator-Centric Active Learning for Subjective NLP Tasks. In EMNLP 2024. [paper]

  • R.X. Lera-Leri, E. Liscio, F. Bistaffa, C.M. Jonker, M. Lopez-Sanchez, P.K. Murukannaiah, J.A. Rodriguez-Aguilar, F. Salas-Molina. 2024. Aggregating Value Systems for Decision Support. In Knowledge-Based Systems, 287, 111453. [paper]

  • M. van der Meer, E. Liscio, C.M. Jonker, A. Plaat, P. Vossen, P.K. Murukannaiah. 2024. A Hybrid Intelligence Method for Argument Mining. In Journal of Artificial Intelligence Research (JAIR). [paper]

2023

  • E. Liscio, O. Araque, L. Gatti, I. Constantinescu, C.M. Jonker, K. Kalimeri, P.K. Murukannaiah. 2023. What does a Text Classifier Learn about Morality? An Explainable Method for Cross-Domain Comparison of Moral Rhetoric. In ACL 2023. [paper] [poster] [slides] [extended abstract]

  • E. Liscio, R. Lera-Leri, F. Bistaffa, R.I.J. Dobbe, C.M. Jonker, M. Lopez-Sanchez, J.A. Rodriguez-Aguilar, P.K. Murukannaiah. 2023. Value Inference in Sociotechnical Systems. In AAMAS ‘23 (position paper). [paper] [poster] [slides] [extended abstract]

2022

  • E. Liscio, A.E. Dondera, A. Geadau, C.M. Jonker, P.K. Murukannaiah. 2022. Cross-Domain Classification of Moral Values. In NAACL 2022 (Findings). [Reproducibility Badge (top 3%)] [paper] [poster] [slide] [video]

  • E. Liscio, M. van der Meer, L.C. Siebert, C.M. Jonker, P.K. Murukannaiah. 2022. What Values should an Agent Align with? An Empirical Comparison of General and Context-Specific Values. In Autonomous Agents and Multi-Agent Systems (JAAMAS). [paper] [poster] [extended abstract] [slide]

  • E. Liscio, C.M. Jonker, P.K. Murukannaiah. 2022. Identifying Context-Specific Values via Hybrid Intelligence. In HHAI2022 (poster track). [Best poster award] [paper] [poster]

  • M. van der Meer, E. Liscio, C.M. Jonker, A. Plaat, P. Vossen, P.K. Murukannaiah. 2022. HyEnA: A Hybrid Method for Extracting Arguments from Opinions. In HHAI2022. [Best paper award] [paper]

  • L.C. Siebert, E. Liscio, P.K. Murukannaiah, L. Kaptein, S. Spruit, J. van den Hoven, C.M. Jonker. 2022. Estimating Value Preferences in a Hybrid Participatory System. In HHAI2022. [Best paper award finalist] [paper]

2021

  • E. Liscio, M. van der Meer, L.C. Siebert, C.M. Jonker, N. Mouter, P.K. Murukannaiah. 2021. Axies: Identifying and Evaluating Context-Specific Values. In AAMAS 2021. [paper] [poster] [slide] [extended abstract] [video]

  • E. Liscio, M. van der Meer, C.M. Jonker, P.K. Murukannaiah. 2021. A Collaborative Platform for Identifying Context-Specific Values. In AAMAS 2021 (demo track). [paper] [video]

Bio

  • 2027–2029 Postdoctoral researcher @ VU Amsterdam (with NWO Veni Grant)
  • 2024–2026 Postdoctoral researcher @ TU Delft (part of AlgoSoc and Hybrid Intelligence)
  • 2020–2024 Cum laude PhD @ TU Delft (part of Hybrid Intelligence)
  • 2017–2020 Deep learning developer and technical project lead at Fizyr
  • 2015–2017 Cum laude MSc. in Systems and Control (TU Delft)
  • 2012–2015 Cum laude BSc. in Automation Engineering (University of Bologna)