Yew-Soon Ong

This paper model MTO as a sequence of sequential transfer optimization problems, concentrating evaluations on a single target per iteration, and proposes a likelihood-informed task prioritization mechanism to maximize transfer utility by identifying the task most likely ready for knowledge integration.

MLLMs can be misled by focusing attacks on phase information, revealing a critical vulnerability that traditional adversarial methods overlook.

PETA achieves superior ligand ranking by adapting pretrained models at test time with only 0.03% of the parameters updated, revolutionizing efficiency in virtual screening.

Evolving adversarial attacks across both text and image modalities can dramatically enhance the transferability and effectiveness of adversarial perturbations in vision-language models.

CVLC achieves a remarkable 16% performance boost in few-shot domain incremental learning, challenging the notion that more data is always necessary for effective adaptation.

VaFM outperforms traditional methods by effectively integrating visual semantics into vehicle routing, addressing complex constraints that were previously overlooked.

Multi-agent RL agents can learn to collaborate *faster* by actively "perceiving" and aligning with each other's policy updates, rather than passively observing environment interactions.

Second-order federated learning can be made robust and practical: FedRCO overcomes instability issues and outperforms first-order methods in non-IID settings.

Scale-PINN slashes PINN training time from hours to minutes by borrowing a key principle from classical numerical solvers: iterative residual correction.