![[2511.07587] Beyond Fact Retrieval: Episodic Memory for RAG with Generative Semantic Workspaces](https://arxiv.org/static/browse/0.3.4/images/arxiv-logo-fb.png){kind=logo}
[2511.07587] Beyond Fact Retrieval: Episodic Memory for RAG with Generative Semantic Workspaces
Summary
The paper presents a novel framework, Generative Semantic Workspace (GSW), designed to enhance long-context reasoning in Large Language Models (LLMs) by integrating episodic memory capabilities, outperforming existing methods in efficiency and performance.
Why It Matters
As LLMs struggle with long-context reasoning, the GSW framework addresses these limitations by providing a structured approach to episodic memory. This advancement is crucial for developing more capable AI agents that can understand and reason over complex narratives, ultimately enhancing AI's applicability in real-world scenarios.
Key Takeaways
- GSW improves long-context reasoning in LLMs by incorporating episodic memory.
- The framework enhances performance by up to 20% on the Episodic Memory Benchmark.
- GSW significantly reduces query-time context tokens by 51%, improving efficiency.
- The approach offers a neuro-inspired method for building coherent narrative representations.
- GSW paves the way for developing AI agents with human-like memory capabilities.
Computer Science > Artificial Intelligence arXiv:2511.07587 (cs) [Submitted on 10 Nov 2025 (v1), last revised 16 Feb 2026 (this version, v2)] Title:Beyond Fact Retrieval: Episodic Memory for RAG with Generative Semantic Workspaces Authors:Shreyas Rajesh, Pavan Holur, Chenda Duan, David Chong, Vwani Roychowdhury View a PDF of the paper titled Beyond Fact Retrieval: Episodic Memory for RAG with Generative Semantic Workspaces, by Shreyas Rajesh and 4 other authors View PDF HTML (experimental) Abstract:Large Language Models (LLMs) face fundamental challenges in long-context reasoning: many documents exceed their finite context windows, while performance on texts that do fit degrades with sequence length, necessitating their augmentation with external memory frameworks. Current solutions, which have evolved from retrieval using semantic embeddings to more sophisticated structured knowledge graphs representations for improved sense-making and associativity, are tailored for fact-based retrieval and fail to build the space-time-anchored narrative representations required for tracking entities through episodic events. To bridge this gap, we propose the \textbf{Generative Semantic Workspace} (GSW), a neuro-inspired generative memory framework that builds structured, interpretable representations of evolving situations, enabling LLMs to reason over evolving roles, actions, and spatiotemporal contexts. Our framework comprises an \textit{Operator}, which maps incoming observations to ...
