Papers by Rohit Saxena
Finding Culture-Sensitive Neurons in Vision-Language Models (2026.eacl-long)
Copied to clipboard
| Challenge: | Vision-language models struggle on culturally situated inputs, study shows . despite impressive performance, many VLMs struggle on such culturally grounded inputs . |
| Approach: | They propose a new margin-based selector to identify neurons associated with cultural selectivity . they also introduce a model-dependent decoder to identify such neurons . |
| Outcome: | The proposed model outperforms probability- and entropy-based methods in identifying neurons associated with cultural selectivity. |
Analyzing LLM Instruction Optimization for Tabular Fact Verification (2026.findings-eacl)
Copied to clipboard
Xiaotang Du, Giwon Hong, Wai-Chung Kwan, Rohit Saxena, Ivan Titov, Pasquale Minervini, Emily Allaway
| Challenge: | evaluating instruction optimization for tabular fact verification is a key challenge for reliable NLP systems. |
| Approach: | They compare instruction optimization for tabular fact verification with a framework based on DSPy . they find that instruction optimization consistently improves verification accuracy . |
| Outcome: | The proposed method improves verification accuracy across four benchmarks and three model families. |
MovieSum: An Abstractive Summarization Dataset for Movie Screenplays (2024.findings-acl)
Copied to clipboard
| Challenge: | Movie screenplay summarization requires an understanding of long input contexts and elements unique to movies. |
| Approach: | They propose a dataset for movie screenplay summarization that includes movie screenplayers accompanied by their Wikipedia plot summaries. |
| Outcome: | The proposed dataset includes 2200 movie screenplays accompanied by their Wikipedia plot summaries. |
Select and Summarize: Scene Saliency for Movie Script Summarization (2024.findings-naacl)
Copied to clipboard
| Challenge: | Existing models for summarizing long-form narrative texts are computationally and memory limited. |
| Approach: | They propose a scene saliency dataset that consists of human-annotated salient scenes for 100 movies. |
| Outcome: | The proposed model outperforms state-of-the-art models and reflects the information content of a movie more accurately than a model that takes the whole movie script as input. |
Are We Done with MMLU? (2025.naacl-long)
Copied to clipboard
Aryo Pradipta Gema, Joshua Ong Jun Leang, Giwon Hong, Alessio Devoto, Alberto Carlo Maria Mancino, Rohit Saxena, Xuanli He, Yu Zhao, Xiaotang Du, Mohammad Reza Ghasemi Madani, Claire Barale, Robert McHardy, Joshua Harris, Jean Kaddour, Emile Van Krieken, Pasquale Minervini
| Challenge: | MMLU is widely adopted but its ground truth errors obscure the true capabilities of LLMs. |
| Approach: | They propose a framework for identifying dataset errors using a novel error annotation protocol and a subset of 5,700 manually re-annotated questions. |
| Outcome: | The proposed framework is based on 5,700 re-annotated questions from the MMLU benchmark. |
What Is That Talk About? A Video-to-Text Summarization Dataset for Scientific Presentations (2025.acl-long)
Copied to clipboard
Dongqi Liu, Chenxi Whitehouse, Xi Yu, Louis Mahon, Rohit Saxena, Zheng Zhao, Yifu Qiu, Mirella Lapata, Vera Demberg
| Challenge: | VISTA dataset contains 18,599 recorded AI conference presentations . large multimodal models exhibit reduced performance in scientific contexts, study shows . |
| Approach: | They propose a dataset specifically designed for video-to-text summarization in scientific domains. |
| Outcome: | This paper compares the performance of large models with human models and shows that they improve on human models. |