視覚言語行動モデルのファインチューニングには、思っているより少ない層で十分である
Finetuning Vision-Language-Action Models Requires Fewer Layers Than You Think
大規模な視覚言語行動（VLA）モデルには層ごとの冗長性があることを発見し、学習不要の構造圧縮パイプラインを提案。冗長な層を除去してモデルを最大50%圧縮し、ファインチューニング時間を40-50%、推論時間を最大30%削減しつつ、性能はフルモデルと同等以上を達成した。
詳しい要約
1. どんなもの？
2. 先行研究と比べてどこがすごい？
3. 技術・手法の肝は？
4. どうやって有効だと検証した？
5. 議論はある？
6. 次に読むべき論文は？
※ AIが要旨から生成した要約です。正確性は原文をご確認ください。
著者: Gia-Binh Nguyen, Trong-Bao Ho, Thien-Loc Ha, Khoa Vo, Philip Lund Møller, Quang T. Nguyen, Long Dinh, Tung M. Luu, Tuan Dam, Vu Duong, Trung Le, Nghi D. Q. Bui, Minh Vu, Tran Nguyen Le, An Thai Le, Ngan Le, Daniel Sonntag, James Zou, Jan Peters, Duy M. H. Nguyen, Ngo Anh Vien
分類: cs.RO, cs.AI
原文アブストラクト
Vision-Language-Action (VLA) models pre-trained on massive video-robot datasets have revolutionized robotic manipulation, yet their multi-billion parameter architectures impose prohibitive computational burdens during downstream fine-tuning and real-time inference. In this work, we reveal a highly non-trivial architectural characteristic of these continuous control foundation policies (e.g., pi_0, GR00T-N1.5): despite being trained on diverse physical trajectories, they exhibit severe layer-wise representational redundancy. To exploit this, we introduce a structural compression pipeline that is entirely training-free, bypassing the need of existing methods to load full-scale models to learn optimized token reductions or dynamic layer selectors. Instead, using only a single forward pass via Centered Kernel Alignment to identify redundant layer features, we remove twin layers to permanently compress the model depth by up to 50% across both the VLM backbone and the continuous control policy head. Downstream fine-tuning of this streamlined architecture yields a dual acceleration benefit: a 40-50% reduction in training time and up to 30% faster real-time inference, while matching or exceeding full-scale base model performance. We comprehensively validate our method across three simulation benchmarks (LIBERO, RoboCasa, SimplerEnv) and 10 diverse real-world manipulation tasks across 4 unique robotic embodiments. These results prove that advanced VLAs require significantly fewer layers than previously assumed, offering a highly compute-efficient paradigm for scalable robot learning.
