Hao Peng

Hao Peng, Yunjia Qi, Xiaozhi Wang et al. · tsinghua

Reinforcement learning with verifiable rewards (RLVR) has become a key technique for enhancing large language models (LLMs), with verification engineering playing a central role. However, best practices for RL in instruction following remain underexplored. In this work, we explore the verification challenge in RL for instruction following and propose VerIF, a verification method that combines rule-based code verification with LLM-based verification from a large reasoning model (e.g., QwQ-32B). To support this approach, we construct a high-quality instruction-following dataset, VerInstruct, containing approximately 22,000 instances with associated verification signals. We apply RL training with VerIF to two models, achieving significant improvements across several representative instruction-following benchmarks. The trained models reach state-of-the-art performance among models of comparable size and generalize well to unseen constraints. We further observe that their general capabilities remain unaffected, suggesting that RL with VerIF can be integrated into existing RL recipes to enhance overall model performance. We have released our datasets, codes, and models to facilitate future research at https://github.com/THU-KEG/VerIF.

Xianghua Zeng, Hao Peng, Dingli Su et al.

Hierarchical Reinforcement Learning (HRL) is a promising approach for managing task complexity across multiple levels of abstraction and accelerating long-horizon agent exploration. However, the effectiveness of hierarchical policies heavily depends on prior knowledge and manual assumptions about skill definitions and task decomposition. In this paper, we propose a novel Structural Information principles-based framework, namely SIDM, for hierarchical Decision Making in both single-agent and multi-agent scenarios. Central to our work is the utilization of structural information embedded in the decision-making process to adaptively and dynamically discover and learn hierarchical policies through environmental abstractions. Specifically, we present an abstraction mechanism that processes historical state-action trajectories to construct abstract representations of states and actions. We define and optimize directed structural entropy, a metric quantifying the uncertainty in transition dynamics between abstract states, to discover skills that capture key transition patterns in RL environments. Building on these findings, we develop a skill-based learning method for single-agent scenarios and a role-based collaboration method for multi-agent scenarios, both of which can flexibly integrate various underlying algorithms for enhanced performance. Extensive evaluations on challenging benchmarks demonstrate that our framework significantly and consistently outperforms state-of-the-art baselines, improving the effectiveness, efficiency, and stability of policy learning by up to 32.70%, 64.86%, and 88.26%, respectively, as measured by average rewards, convergence timesteps, and standard deviations.

Qi Wu, Yingguang Yang, hao liu et al.

Social bot detection is crucial for mitigating misinformation, online manipulation, and coordinated inauthentic behavior. While existing neural network-based detectors perform well on benchmarks, they struggle with generalization due to distribution shifts across datasets and frequently produce overconfident predictions for out-of-distribution accounts beyond the training data. To address this, we introduce a novel Uncertainty Estimation for Social Bot Detection (UESBD) framework, which quantifies the predictive uncertainty of detectors beyond mere classification. For this task, we propose Robust Multi-modal Neural Processes (RMNP), which aims to enhance the robustness of multi-modal neural processes to modality inconsistencies caused by social bot camouflage. RMNP first learns unimodal representations through modality-specific encoders. Then, unimodal attentive neural processes are employed to encode the Gaussian distribution of unimodal latent variables. Furthermore, to avoid social bots stealing human features to camouflage themselves thus causing certain modalities to provide conflictive information, we introduce an evidential gating network to explicitly model the reliability of modalities. The joint latent distribution is learned through the generalized product of experts, which takes the reliability of each modality into consideration during fusion. The final prediction is obtained through Monte Carlo sampling of the joint latent distribution followed by a decoder. Experiments on three real-world benchmarks show the effectiveness of RMNP in classification and uncertainty estimation, as well as its robustness to modality conflicts.

Hao Peng, Xiaoli Bai

Due to the lack of information such as the space environment condition and resident space objects' (RSOs') body characteristics, current orbit predictions that are solely grounded on physics-based models may fail to achieve required accuracy for collision avoidance and have led to satellite collisions already. This paper presents a methodology to predict RSOs' trajectories with higher accuracy than that of the current methods. Inspired by the machine learning (ML) theory through which the models are learned based on large amounts of observed data and the prediction is conducted without explicitly modeling space objects and space environment, the proposed ML approach integrates physics-based orbit prediction algorithms with a learning-based process that focuses on reducing the prediction errors. Using a simulation-based space catalog environment as the test bed, the paper demonstrates three types of generalization capability for the proposed ML approach: 1) the ML model can be used to improve the same RSO's orbit information that is not available during the learning process but shares the same time interval as the training data; 2) the ML model can be used to improve predictions of the same RSO at future epochs; and 3) the ML model based on a RSO can be applied to other RSOs that share some common features.