AMTA 2025
Emerging AI Breakthroughs and Challenges in Translation Automation
Location
- Online
Links
Schedule
| 9:30 – 10:00 | Networking | Room 1 |
| 10:00 – 10:15 | Welcome Jay Marciano | Room 1 |
| 10:15 – 11:00 | KEYNOTE — The Multilingual Nexus: Exploring the Intersection of Translation and LLMs Julia Kreutzer | Room 1 |
| 11:10 – 11:30 | Semi-Synthetic Parallel Data for Translation Quality Estimation Ilan Kernerman | Room 1 |
| 11:10 – 11:30 | From Neural MT to Generative AI Translation: Evolving Training Paradigms for Cognitive Load Reduction Viveta Gene | Room 2 |
| 11:10 – 11:30 | Scoring the Match: MQM-Based Quality Evaluation of LLM and NMT for Football Translation Huiting Zeng | Room 3 |
| 11:35 – 11:55 | Grounding Real-Time Speech Translation Through Visual Cues Claudio Fantinuoli | Room 1 |
| 11:35 – 11:55 | Navigating the Future of Linguistic Talent Mara Nunziatini | Room 2 |
| 11:35 – 11:55 | Enhancing Human Linguistic Quality Assurance with LLMs: Comprehensive and Client-Specific Translation Evaluation Olesia Khrapunova | Room 3 |
| 12:00 – 12:20 | Enhancing Machine Translation through Multi-Agent Communication: A Post-Editing Approach Marina Albert Girona | Room 1 |
| 12:00 – 12:20 | Gender Bias, Anaphora, and Pragmatics: A Disentanglement Study in Machine Translation Celia Soler | Room 2 |
| 12:00 – 12:20 | Leveraging Knowledge Graphs for Translation Marina Sánchez Torrón | Room 3 |
| 12:25 – 12:45 | Utilizing AutoMQM to automatically correct the translation through Self-improvement and Reward Augmentation James Lin | Room 1 |
| 12:25 – 12:45 | Does AI Have a Style? A Study on Literary Authorship and Translation Natalia Resende | Room 2 |
| 12:25 – 12:45 | Translation Quality and Error Annotation in Legal Translation: A Comparative Analysis of LLMs and NMT Systems Stephanie Rodriguez | Room 3 |
| 12:45 – 1:15 | 🍴 | |
| 1:15 – 2:00 | PANEL | Room 1 |
| 2:05 – 2:25 | Fine-tuning EuroLLM for Specialized Domains: A Practical Guide to Deployable Domain Models Maciej Modrzejewski | Room 1 |
| 2:05 – 2:25 | How to build a bomb and other tales: How red teaming a large language model improves safety Patricia Pandeiro | Room 2 |
| 2:05 – 2:25 | Introducing PAEM-CMT: Purpose-Aligned Evaluation Metric of Customized Machine Translation Paşa Bayramoğlu | Room 3 |
| 2:30 – 2:50 | General to Specialised: An Adaptive Framework for Domain-specific Translation Leonor Graça | Room 1 |
| 2:30 – 2:50 | Time to switch gears? Analysis of an NMT-first to LLM-only workflow for localization Julian Hamm | Room 2 |
| 2:55 – 3:15 | Leveraging Large Language Models for Pharmaceutical Documents Andreia Lopes | Room 1 |
| 2:55 – 3:15 | A Modular Framework for Intelligent MT Engine Selection Leveraging Generative AI and Hybrid Metrics Inacio Vieira | Room 2 |
| 2:55 – 3:15 | Art and craft: the reality of designing practical applications of AI in the localization industry Carla Schelfhout | Room 3 |
| 3:20 – 3:40 | Terminology Augmented Generation: An Agentic Approach to Optimize LLM Translation Output by Supplying Terminology and Metadata via MCP Klaus Fleischmann | Room 1 |
| 3:20 – 3:40 | When Not to Use LLMs Alex Yanishevsky | Room 2 |
| 3:40 – 3:55 | Coffee Networking | Room 1 |
| 4:00 – 4:20 | The Rise of LLM MT and the Changing MT User Experience Kirti Vashee | Room 1 |
| 4:00 – 4:20 | Translating Trust: Dow Jones’s Journey with AI Translation Konstantin Savenkov | Room 2 |
| 4:00 – 4:20 | Walk through translation AI options with the Machine Translation Foundation Adam Bittlingmeyer | Room 3 |
| 4:25 – 4:45 | On the Origins of the Translation Abilities of Large Language Models Yuri Balashov | Room 1 |
| 4:25 – 4:45 | Measuring What Matters: A Multilingual Framework for Real-World LLM Evaluation David Harper | Room 2 |
| 4:50 – 5:10 | Cognitive Reasoning in Translation: Evaluating Chain-of-Thought, Explaining, Metacognition, and Critique in Humans and General-Purpose vs. Advanced-Reasoning LLMs Ming Qian | Room 1 |
| 4:50 – 5:10 | State of Translation Automation 2025 Daria Sinitsyna | Room 2 |
| 4:50 – 5:10 | Leveraging AI Tools in Interpreting: Demonstrations, Insights, and Research David Proano Celi | Room 3 |
| 5:15 – 5:35 | Empowering Reliable Use of Machine Translation through Q&A Feedback Ki Dayeon | Room 1 |
| 5:40 – 5:50 | Closing Jay Marciano | Room 1 |
| 5:50 – 6:30 | Afterhours Networking | Room 1 |