From Supervision to Exploration: What Does Protein Language Model Learn During Reinforcement Learning?

Protein language models (PLMs) have advanced computational protein science through large-scale pretraining and scalable architectures. In parallel, reinforcement learning (RL) has broadened exploration and enabled precise multi-objective optimization in protein design. Yet whether RL can push PLMs beyond their pretraining priors to uncover latent sequence-structure-function rules remains unclear. We address this by pairing RL with PLMs across four domains: antimicrobial peptide design, kinase variant optimization, antibody engineering, and inverse folding. Using diverse RL algorithms and model classes, we ask if RL improves sampling efficiency and, more importantly, if it reveals capabilities not captured by supervised learning. Across benchmarks, RL consistently boosts success rates and sample efficiency. Performance follows a three-factor interaction: task headroom, reward fidelity, and policy capacity jointly determine gains. When rewards are accurate and informative, policies have sufficient capacity, and tasks leave room beyond supervised baselines, improvements scale; when rewards are noisy or capacity is constrained, gains saturate despite exploration. This view yields practical guidance for RL in protein design: prioritize reward modeling and calibration before scaling policy size, match algorithm and regularization strength to task difficulty, and allocate capacity where marginal gains are largest. Implementation is available at https://github.com/chq1155/RL-PLM.

Gnothi Seauton: Empowering Faithful Self-Interpretability in Black-Box Models

The debate between self-interpretable models and post-hoc explanations for black-box models is central to Explainable AI (XAI). Self-interpretable models, such as concept-based networks, offer insights by connecting decisions to human-understandable concepts but often struggle with performance and scalability. Conversely, post-hoc methods like Shapley values, while theoretically robust, are computationally expensive and resource-intensive. To bridge the gap between these two lines of research, we propose a novel method that combines their strengths, providing theoretically guaranteed self-interpretability for black-box models without compromising prediction accuracy. Specifically, we introduce a parameter-efficient pipeline, *AutoGnothi*, which integrates a small side network into the black-box model, allowing it to generate Shapley value explanations without changing the original network parameters. This side-tuning approach significantly reduces memory, training, and inference costs, outperforming traditional parameter-efficient methods, where full fine-tuning serves as the optimal baseline. *AutoGnothi* enables the black-box model to predict and explain its predictions with minimal overhead. Extensive experiments show that *AutoGnothi* offers accurate explanations for both vision and language tasks, delivering superior computational efficiency with comparable interpretability.

Robust Trajectory and Offloading for Energy-Efficient UAV Edge Computing in Industrial Internet of Things

Efficient data processing and computation are essential for the industrial Internet of things (IIoT) to empower various applications, which yet can be significantly bottlenecked by the limited energy capacity and computation capability of the IIoT nodes. In this paper, we employ an unmanned aerial vehicle (UAV) as an edge server to assist IIoT data processing, while considering the practical issue of UAV jittering. Specifically, we propose a joint design on trajectory and offloading strategies to minimize energy consumption due to local and edge computation, as well as data transmission. We particularly address the UAV jittering that induces Gaussian-distributed uncertainties associated with flying waypoints, resulting in probabilistic-form flying speed and data offloading constraints. We exploit the Bernstein-type inequality to reformulate the constraints in deterministic forms and decompose the energy minimization to solve for trajectory and offloading separately within an alternating optimization framework. The subproblems are then tackled with the successive convex approximation technique. Simulation results show that our proposal strictly guarantees robustness under uncertainties and effectively reduces energy consumption as compared with the baselines.