Be Confident: Uncovering Overfitting in MLLM Multi-Task Tuning

Wenke Huang, Jian Liang, Guancheng Wan, Didi Zhu, He Li +4 more
2/14/2026

No official code found. Best implementation: WenkeHuang/NRCA (2 stars, last commit 2025-09-24).

Semantic Scholar

Code Implementations(1)

WenkeHuang/NRCACommunity72%

ICML 2025 - Be Confident: Uncovering Overfitting in MLLM Multi-Task Tuning

20CSS, HTMLSep 24, 20251 years ago
mllm

Cite this paper

@article{huang2026confident,
  title  = {Be Confident: Uncovering Overfitting in MLLM Multi-Task Tuning},
  author = {Wenke Huang and Jian Liang and Guancheng Wan and Didi Zhu and He Li and Jiawei Shao and Mang Ye and Bo Du and Dacheng Tao},
  year   = {2026},
  url    = {https://api.semanticscholar.org/CorpusID:6f7edd6913f4b4920f59db5ba5b85654694ba997},
  journal = {ICML 2025 2025}
}

Discussion