data contamination
Data contamination occurs when training data inadvertently includes unwanted information or labels, leading to biased or incorrect model outputs and undermining model validity.
- A Controllable Examination for Long-Context Language Models
- AudSemThinker: Enhancing Audio-Language Models Through Reasoning over Semantics of Sound
- Generating Computational Cognitive models using Large Language Models
- LiveCodeBench Pro: How Do Olympiad Medalists Judge LLMs in Competitive Programming?
- PHYBench: Holistic Evaluation of Physical Perception and Reasoning in Large Language Models
- Position: Benchmarking is Broken - Don't Let AI be Its Own Judge
- STEER-ME: Assessing the Microeconomic Reasoning of Large Language Models
- The Illusion of Thinking: Understanding the Strengths and Limitations of Reasoning Models via the Lens of Problem Complexity
- ThinkBench: Dynamic Out-of-Distribution Evaluation for Robust LLM Reasoning