ParamMem: Augmenting Language Agents with Parametric Reflective Memory
This work is significant for researchers and practitioners working on language agents, as it provides a method to enhance reasoning performance by improving the diversity of self-reflection, which is an incremental improvement to existing reflection-based agents.
This paper addresses the problem of repetitive self-reflection in language agents, which limits reasoning performance. They introduce ParamMem, a parametric memory module that encodes cross-sample reflection patterns, enabling diverse reflection generation. This approach consistently improves performance over state-of-the-art baselines in code generation, mathematical reasoning, and multi-hop question answering.
Self-reflection enables language agents to iteratively refine solutions, yet often produces repetitive outputs that limit reasoning performance. Recent studies have attempted to address this limitation through various approaches, among which increasing reflective diversity has shown promise. Our empirical analysis reveals a strong positive correlation between reflective diversity and task success, further motivating the need for diverse reflection signals. We introduce ParamMem, a parametric memory module that encodes cross-sample reflection patterns into model parameters, enabling diverse reflection generation through temperature-controlled sampling. Building on this module, we propose ParamAgent, a reflection-based agent framework that integrates parametric memory with episodic and cross-sample memory. Extensive experiments on code generation, mathematical reasoning, and multi-hop question answering demonstrate consistent improvements over state-of-the-art baselines. Further analysis reveals that ParamMem is sample-efficient, enables weak-to-strong transfer across model scales, and supports self-improvement without reliance on stronger external model, highlighting the potential of ParamMem as an effective component for enhancing language agents.