Newest papers

WeAct: Dynamic Weight Switching as Agent Actions

Rui Shi, Xunjian Yin, Jiacheng Sang, Zichuan Liu, Shuyan Zhou · 2026 · arXiv (Cornell University)

DOI: 10.48550/arxiv.2610.26847 · cited by 0 works (OpenAlex) · open-access PDF

French

Les agents à horizon long doivent apprendre des capacités hétérogènes à partir de trajectoires interliées. Mettre à jour les paramètres partagés pour apprendre de nouvelles capacités peut dégrader celles acquises auparavant, entraînant l'oubli catastrophique. À mesure qu'une trajectoire se déploie, les différentes étapes peuvent nécessiter des capacités différentes, mais la formation conjointe conventionnelle et la fusion d'adaptateurs statiques maintiennent une configuration paramétrique fixe tout au long de l'exécution. Nous introduisons WeAct (Actions internes de poids), qui entraîne des adaptateurs LoRA indépendants spécifiques aux capacités sur une colonne vertébrale gelée et ajoute une sélection d'adaptateur à l'espace d'action

Machine translation (Meta M2M100); the English original follows.

Abstract

Long-horizon agents must learn heterogeneous capabilities from interleaved trajectories. Updating shared parameters to learn new capabilities can degrade previously acquired ones, resulting in catastrophic forgetting. As a trajectory unfolds, different stages may require different capabilities, yet conventional joint training and static adapter merging maintain a fixed parameter configuration throughout execution. We introduce WeAct (Weight-level Internal Actions), which trains independent capability-specific LoRA adapters over a frozen backbone and adds adapter selection to the agent's action space. These internal actions use the model's native tool-calling interface for on-demand capability composition in a shared context. Across eight deep-research benchmarks, Qwen3-32B with WeAct improves average accuracy by 6.2% relative to full fine-tuning counterparts despite using less compute, and by 27.5% relative to the strongest static LoRA-merging method. WeAct also supports modular continual adaptation, by adding a new code computation adapter to expand the agent's capabilities and updating the existing writing adapter to improve report quality. Both updates leave the backbone and other adapters unchanged, supporting modular continual adaptation in agents.

Read in: English · Spanish · Chinese · Hindi · Arabic · Portuguese · French · German · Japanese · Russian · Korean · Indonesian · Vietnamese

Topics: Adapter (computing), Modular design, Computer science, Train, Computation, Distributed computing

Sources: OpenAlex (CC0), arXiv, Crossref. Agent Exchange