Scholar
Ruchira Dhar
Google Scholar ID: YgAMn9gAAAAJ
PhD Researcher, University of Copenhagen
Citations & Impact
Citations
19
H-index
2
i10-index
1
Publications
8
Co-authors
7
list available
Publications
Resume (English only)
Academic Achievements
- Paper 'Beyond Technocratic XAI: The Who, What & How in Explanation Design', AIES 2025 (Main Conference)
- Paper 'Mechanistic Interpretability Needs Philosophy', under review
- Paper 'Evaluating Multimodal Language Models as Visual Assistants for Visually Impaired Users', ACL 2025 (Main Conference)
- Paper 'Investigating the Role of Modality and Training Objective on Representational Alignment Between Transformers and the Brain', NeurIPS 2024 (UniReps Workshop)
- Paper 'Defining Knowledge: Bridging Epistemology and Large Language Models', EMNLP 2024 (Main Conference)
- Presentation 'From Words to Worlds: Compositionality for Cognitive Architectures', LLMs & Cognition Workshop @ ICML 2024
- Presentation 'The Completeness Problem: Beyond Human Metrics in Assessing Abilities of Cognitive Systems', ACAIN 2024
- Presentation 'Compositionality in Language Models: A Perspective in Changing Interpretations and Methods', AIAI 2024
Background
- PhD Researcher in Natural Language Processing and AI
- Research focuses on evaluation of Large Language Models (LLMs): How to determine if a model is 'good' or 'safe', whether current evaluation practices are rigorous enough, and how to make model strengths and limitations more interpretable and actionable for the public
- Interested in AI Policy and Governance, especially the role of evaluation practices
- Affiliated with the CoAStaL Lab, Department of Computer Science, University of Copenhagen, supervised by Dr. Anders Søgaard
- Also involved with the Centre of Philosophy of AI, exploring societal and ethical dimensions of AI
