Papers by Irina Rish
Training Dynamics Underlying Language Model Scaling Laws: Loss Deceleration and Zero-Sum Learning (2025.acl-long)
Copied to clipboard
| Challenge: | Increasing language model size improves cross-entropy loss with power-law behaviour, but scaling laws do not explain how scaling improves loss. |
| Approach: | They find that language models undergo loss deceleration early in training . they attribute loss deceleration to a type of degenerate training dynamics we call zero-sum learning . |
| Outcome: | The proposed scaling improves loss on language models, but degrades loss in other subsets, resulting in bottlenecks. |
Improving Adversarial Robustness in Vision-Language Models with Architecture and Prompt Design (2024.findings-emnlp)
Copied to clipboard
| Challenge: | Vision-Language Models (VLMs) have seen a significant increase in research interest and real-world applications, including healthcare, autonomous systems, and security. |
| Approach: | They propose novel approaches to enhance model robustness through prompt engineering by suggesting adversarial perturbations or rephrasing questions. |
| Outcome: | The proposed approaches improve model robustness against strong image-based attacks such as Auto-PGD. |
Scaling Laws and Efficient Inference for Ternary Language Models (2025.acl-long)
Copied to clipboard
Tejas Vaidhya, Ayush Kaushal, Vineet Jain, Francis Couture-Harpin, Prashant Shishodia, Majid Behbahani, Yuriy Nevmyvaka, Irina Rish
| Challenge: | Large language models (LLMs) are increasingly used across research and industry applications, yet their inference efficiency remains a challenge. |
| Approach: | They propose ternary language models that employ quantization-aware training to significantly reduce memory requirements. |
| Outcome: | The proposed ternary language models demonstrate sustained performance gains at scale. |
GitChameleon 2.0: Evaluating AI Code Generation Against Python Library Version Incompatibilities (2026.acl-long)
Copied to clipboard
Diganta Misra, Nizar Islah, Victor May, Brice Rauby, Zihan Wang, Justine Gehring, Antonio Orvieto, Muawiz Sajjad Chaudhary, Eilif B. Muller, Irina Rish, Samira Ebrahimi Kahou, Massimo Caccia
| Challenge: | Existing code evolution benchmarks lack execution-based evaluation for generating code compliant with specific library versions. |
| Approach: | They propose a new Python code completion problem that evaluates the ability of large language models to perform version-conditioned code generation. |
| Outcome: | The proposed benchmarks show that state-of-the-art systems can perform version-conditioned code generation with high success rates. |
CAVE : Detecting and Explaining Commonsense Anomalies in Visual Environments (2025.emnlp-main)
Copied to clipboard
Rishika Bhagwatkar, Syrielle Montariol, Angelika Romanou, Beatriz Borges, Irina Rish, Antoine Bosselut
| Challenge: | a new benchmark for computer vision fails to capture richness and unpredictability of real-world anomalies . state-of-the-art VLMs struggle with visual anomaly perception and commonsense reasoning . elucidating the nature of anomalies is a fundamental human trait . |
| Approach: | They propose a benchmark for visual anomalies that includes annotations for visual grounding and categorizing anomalies based on their visual manifestations, their complexity, severity, and commonness. |
| Outcome: | The proposed benchmark improves on existing vision models by incorporating visual annotations. |