世界モデルのためのSenseパラダイム
The Sense Paradigm for World Models
埋め込み空間の距離に意味を持たせる単一のクロスエントロピー損失「Sense」を提案し、不可逆遷移の検出と計画精度で既存手法を大きく上回った。
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: Haidong Hu
分類: cs.LG
原文アブストラクト
Latent world models predict the future in an embedding space rather than in pixels. An encoder trained only to predict collapses and maps every observation to the same point, so existing models add a mechanism against collapse, such as a target network, a variance penalty or a distribution penalty. These mechanisms keep the embedding from collapsing, but they leave its distances without meaning, and the model cannot tell a reversible transition from an irreversible one. We propose Sense, a single cross-entropy loss over the directed distance $d(x\to y)=\sum_i \mathrm{ReLU}(y_i-x_i)$ that replaces these mechanisms. The loss avoids collapse on its own, and since the encoder no longer needs the predictor, the predictor is fitted afterwards to fixed embeddings by plain regression. On four pixel environments, the return distance learned by Sense detects irreversible transitions with an AUC between 0.95 and 1.00, where LeWorldModel stays between 0.39 and 0.59. On TwoRoom, the navigation benchmark of LeWorldModel, planning with Sense reaches the goal in 100% of episodes against 91% for LeWorldModel. As the goal moves farther away, the planning failure rate of every baseline grows, while that of Sense stays near zero in Cup, Cliff and Sokoban and is the lowest of all methods in Pendulum.
