compositional reasoning
This involves the AI's ability to combine different pieces of information or logic constructs to form coherent and meaningful outputs, critical for solving complex problems that require integrative thinking.
- Advancing Compositional Awareness in CLIP with Efficient Fine-Tuning
- CF-VLM：CounterFactual Vision-Language Fine-tuning
- Enhancing Compositional Reasoning in CLIP via Reconstruction and Alignment of Text Descriptions
- Ineq-Comp: Benchmarking Human-Intuitive Compositional Reasoning in Automated Theorem Proving of Inequalities
- MMPerspective: Do MLLMs Understand Perspective? A Comprehensive Benchmark for Perspective Perception, Reasoning, and Robustness
- MemEIC: A Step Toward Continual and Compositional Knowledge Editing
- OMEGA: Can LLMs Reason Outside the Box in Math? Evaluating Exploratory, Compositional, and Transformative Generalization
- PARTONOMY: Large Multimodal Models with Part-Level Visual Understanding
- Visual Structures Help Visual Reasoning: Addressing the Binding Problem in LVLMs