Reproduce “YANAMI: A Large-Scale Multimodal Dataset and Benchmark for Multiturn Stateful Interaction” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Shengyu Wang, Tengyue Nong, Junming Liu, Haosong Peng, Siyu Li, Ruiqi Lei, Fushuo Huo, Yuhao Pan and 5 more
YANAMI constructs multi-turn visual dialogues in which later requests depend on accumulated constraints, providing training data and a benchmark for stateful image interaction. Its 402,639 dialogues train models, while YANAMI-Bench evaluates 400 dialogues across four scenarios. Fine-tuning 7B BAGEL on YANAMI-400K raises its score from 65.02 to 78.25, a 13.23-point gain, but does not isolate the effect of state supervision.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.