Test-Time Scaling (TTS) is a promising approach to progressively elicit the model’s intelligence during inference. Recently, training-based TTS methods, such as continued reinforcement learning (RL), have further surged in popularity, while training-free TTS methods are gradually fading from prominence. However, the additional computation overhead of training amplifies the burden on test-time scaling.In this paper, we focus on training-free TTS methods for reasoning. We first design Conditional Step-level Self-refinement, a fine-grained sequential scaling method guided by process verification. On top of its effectiveness, we further combine it with other classical parallel scaling methods at the step level, to introduce a novel inference paradigm called Hybrid Test-Time Scaling. Extensive experiments on five instruction-tuned LLMs across different scales (3B-14B) and families demonstrate that hybrid strategy incorporating various training-free TTS methods at a fine granularity has considerable potential for expanding the reasoning performance boundaries of LLMs.

Social network simulation is developed to provide a comprehensive understanding of social networks in the real world, which can be leveraged for a wide range of applications such as group behavior emergence, policy optimization, and business strategy development. However, billions of individuals and their evolving interactions involved in social networks pose challenges in accurately reflecting real-world complexities. In this study, we propose a comprehensive Social network Simulation System (GA-S³) that leverages newly designed Group Agents to make intelligent decisions regarding various online events. Unlike other intelligent agents that represent an individual entity, our group agents model a collection of individuals exhibiting similar behaviors, facilitating the simulation of large-scale network phenomena with complex interactions at a manageable computational cost. Additionally, we have constructed a social network benchmark from 2024 popular online events that contains fine-grained information on Internet traffic variations. The experiment demonstrates that our approach is capable of achieving accurate and highly realistic prediction results.

“Reinforcement learning with human feedback for aligning large language models (LLMs) trainsa reward model typically using ranking loss with comparison pairs. However, the training pro-cedure suffers from an inherent problem: the uncontrolled scaling of reward scores during rein-forcement learning due to the lack of constraints while training the reward model. This paperproposes a Prior Constraints-based Reward Model (PCRM) training method to mitigate thisproblem. PCRM incorporates prior constraints—specifically, length ratio and cosine similaritybetween outputs of each comparison pair—during reward model training to regulate optimiza-tion magnitude and control score margins. We comprehensively evaluate PCRM by examining itsrank correlation with human preferences and its effectiveness in aligning LLMs via RL. Exper-imental results demonstrate that PCRM significantly improves alignment performance by effec-tively constraining reward score scaling. As another bonus, our method is easily integrated intoarbitrary rank-based alignment methods, such as direct preference optimization, and can yieldconsistent improvement. The code is available at https://github.com/wangclnlp/DeepSpeed-Chat-Extension/tree/PCRM.”

Alignment training is crucial for enabling large language models (LLMs) to cater to human intentions and preferences. It is typically performed based on two stages with different objectives: instruction-following alignment and human-preference alignment. However, aligning LLMs with these objectives in sequence suffers from an inherent problem: the objectives may conflict, and the LLMs cannot guarantee to simultaneously align with the instructions and human preferences well. To response to these, in this work, we propose a Hybrid Alignment Training (Hbat) approach, based on alternating alignment and modified elastic weight consolidation methods. The basic idea is to alternate between different objectives during alignment training, so that better collaboration can be achieved between the two alignment tasks. We experiment with Hbat on summarization and dialogue tasks. Experimental results show that the proposed Hbat can significantly outperform all baselines. Notably, Hbat yields consistent performance gains over the traditional two-stage alignment training when using both proximal policy optimization and direct preference optimization.