Publications and Manuscripts

When AI Benchmarks Plateau: A Systematic Study of Benchmark Saturation

Mubashara Akhtar,  Anka Reuel,  Prajna Soni,  Sanchit Ahuja,  Pawan Sasanka Ammanamanchi,  Ruchit Rawal,  Vilém Zouhar,  Srishti Yadav,  Chenxi Whitehouse,  Dayeon Ki,  Jennifer Mickel,  Leshem Choshen,  Marek Šuppa,  Jan Batzner,  Jenny Chim,  Jeba Sania,  Yanan Long,  Hossein A. Rahmani,  Christina Knight,  Yiyang Nan,  Jyoutir Raj,  Yu Fan,  Shubham Singh,  Subramanyam Sahoo,  Eliya Habba,  Usman Gohar,  Siddhesh Pawar,  Robert Scholz,  Arjun Subramonian,  Jingwei Ni,  Mykel Kochenderfer,  Sanmi Koyejo,  Mrinmaya Sachan,  Stella Biderman,  Zeerak Talat,  Avijit Ghosh,  Irene Solaiman

ICML, 2026

arXiv - project

Aya Vision: Advancing the Frontier of Multilingual Multimodality

Saurabh Dash*,  Yiyang Nan*,  John Dang,  Arash Ahmadian,  Shivalika Singh,  Madeline Smith,  Bharat Venkitesh,  Vlad Shmyhlo,  Viraat Aryabumi,  Walter Beller-Morales,  Jeremy Pekmez,  Jason Ozuzu,  Pierre Richemond,  Acyr Locatelli,  Nick Frosst,  Phil Blunsom,  Aidan Gomez,  Ivan Zhang,  Marzieh Fadaee,  Manoj Govindassamy,  Sudip Roy,  Matthias Gallé,  Beyza Ermis,  Ahmet Üstün,  Sara Hooker

arXiv - blogpost - huggingface

The Leaderboard Illusion

Shivalika Singh,  Yiyang Nan,  Alex Wang,  Daniel D'Souza,  Sayash Kapoor,  Ahmet Üstün,  Sanmi Koyejo,  Yuntian Deng,  Shayne Longpre,  Noah Smith,  Beyza Ermis,  Marzieh Fadaee,  Sara Hooker

NeurIPS Datasets and Benchmarks Track, 2025

arXiv - blogpost - media coverage: New Scientist

INCLUDE: Evaluating Multilingual Language Understanding with Regional Knowledge

Angelika Romanou,  Negar Foroutan,  Anna Sotnikova,  Sree Harsha Nelaturu,  Shivalika Singh,  Rishabh Maheshwary,  Micol Altomare,  Zeming Chen,  Mohamed A. Haggag,  ...Yiyang Nan (with 45 other authors)...,  Imanol Schlag,  Marzieh Fadaee,  Sara Hooker,  Antoine Bosselut

ICLR, 2025 (Spotlight)

arXiv - dataset

Learning to Generate Instruction Tuning Datasets for Zero-Shot Task Adaptation

Nihal V. Nayak,  Yiyang Nan,  Avi Trost,  Stephen H. Bach

ACL Findings, 2024

arXiv - code - huggingface - Instruction Tuning Workshop @ NeurIPS, 2023

Hear Me, See Me, Understand Me: Audio-Visual Autism Behavior Recognition

Shijian Deng,  Erin E. Kosloski,  Siddhi Patel,  Zeke Aharon Barnett,  Yiyang Nan,  Alexander Kaplan,  Sisira Aarukapalli,  William T. Doan,  Matthew Wang,  Harsh Singh,  Pamela R Rollins,  Yapeng Tian

IEEE Transactions on Multimedia, 2024

arXiv - dataset - Sight and Sound Workshop @ CVPR, 2024

SaSR-Net: Source-Aware Semantic Representation Network for Enhancing Audio-Visual Question Answering

Tianyu Yang,  Yiyang Nan,  Lisen Dai,  Zhenwen Liang,  Yapeng Tian,  Xiangliang Zhang

EMNLP Findings, 2024

arXiv

Continual Audio-Visual Sound Separation

Weiguo Pian,  Yiyang Nan,  Shijian Deng,  Shentong Mo,  Yunhui Guo,  Yapeng Tian

NeurIPS, 2024

arXiv - code

Harmonizing across datasets to improve the transferability of drug combination prediction

Hanrui Zhang*,  Ziyan Wang*,  Yiyang Nan*,  Bulat Zagidullin*,  Daiyao Yi,  Jing Tang,  Yuanfang Guan

Communications Biology, 2023

PubMed