Evaluate dataset discovery and synthesis capabilities for text summarization with political news articles.

Assess multilingual dataset construction for English-Tamil medical translation tasks.

Examine knowledge base summarization through document-level narrative text processing.

Investigate medical question answering using USMLE-style multiple choice clinical scenarios.

Develop systematic web data cleaning methodologies for large-scale corpus preprocessing.

Design mathematical problem curation strategies for reasoning model training optimization.

Implement contamination detection and difficulty filtering for code instruction datasets.

Enhance multidisciplinary scientific reasoning through advanced fine-tuning techniques.

Advance mathematical problem-solving via sophisticated training methodologies.

Synthesize search-augmented reasoning data for supervised fine-tuning applications.

Generate temporal Theory of Mind scenarios for social interaction understanding.

Augment scientific visual reasoning through multimodal training enhancement.

Develop efficient model realignment algorithms for training-efficient adjustment.

Implement entropy collapse prevention strategies in reinforcement learning training.

Design robust preference optimization methods for noisy preference data.

Create reward functions for search-augmented reasoning in reinforcement learning.

Develop unified reward functions for multi-platform GUI grounding models.

Build prompt-based deep research agents using foundation model orchestration.

Construct efficient mathematical reasoning workflows for complex problem solving.

Develop visual reasoning systems for spatial and semantic understanding tasks.