LLM reasoning / chain-of-thought
TVI-CoT: Text-Visual Interleaved Chain-of-Thought Reasoning for Multimodal Understanding
Current frontier — recent, not yet superseded in the knowledge base
0 papers critique it · 0 beat it on benchmarks
Recent methods in the same sub-problem, not yet superseded in the knowledge base — arXiv benchmark leaders, not vetted production recommendations.