Five Algoverse research teams have been accepted to workshops at ICLR 2025 in Singapore. ICLR is one of the most prestigious conferences in deep learning.
Accepted papers:
1. MALIBU Benchmark: Multi-Agent LLM Implicit Bias Uncovered Authors: Imran Mirza, Cole Huang, Ishwara Vasista, Rohan Patil Venue: Building Trust in LLMs @ ICLR 2025; NAACL SRW 2025
2. EnDive: A Cross-Dialect Benchmark for Fairness and Performance in Large Language Models Authors: Abhay Gupta, Jacob Cheung, Philip Meng, Shayan Sayyed Venue: Building Trust in LLMs @ ICLR 2025; NAACL SRW 2025; EMNLP Findings 2025
3. Finding Sparse Autoencoder Representations Of Errors In CoT Prompting Authors: Justin Theodorus, V Swaytha, Shivani Gautam, Adam Ward, Mahir Shah Venue: Building Trust in LLMs @ ICLR 2025
4. Deconstructing Bias: A Multifaceted Framework for Diagnosing Cultural and Compositional Inequities in Text-to-Image Generative Models Authors: Muna Said, Aarib Haider, Rabia Usman, Sonia Okon Venue: SynthData @ ICLR 2025
5. Efficient Transformers via MPO-Based Low-Rank Factorization and Pruning Authors: Sam Mikhak, Venkata Sai Gummidi Venue: Sparsity in LLMs @ ICLR 2025
