Research map / Large language models / Retrieval and knowledge
Beyond Fact Retrieval: Episodic Memory for RAG with Generative Semantic Workspaces
Shreyas Rajesh, Pavan Holur, Chenda Duan, David Chong, Vwani Roychowdhury
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| replace chunk/graph retrieval with GSW structured episodic memory | GPT-4o QA · EpBench-200 overall | F1: 0.771 -> 0.850 | Improved |
| use GSW instead of HippoRAG2 on multi-cue queries | GPT-4o QA · EpBench-200 6+ cues | recall: 0.675 -> 0.822 | Improved |
| use GSW at 10x corpus scale | GPT-4o QA · EpBench-2000 overall | F1: 0.675 -> 0.773 | Improved |
| use entity-specific GSW summaries rather than chunks | GPT-4o QA · EpBench-200 | context tokens: 8771 -> 3587 (vs embedding RAG); 51% fewer than GraphRAG | Improved |
| GSW vs GraphRAG on queries with one cue | GPT-4o QA · EpBench-200 1 cue | F1: 0.625 -> 0.744 | Improved |
| GSW vs Embedding RAG on 1-cue precision | GPT-4o QA · EpBench-200 1 cue | precision: 0.745 -> 0.755 | No clear effect |
| GSW vs HippoRAG2 6+ cue precision | GPT-4o QA · EpBench-200 6+ cues | precision: 0.940 -> 0.890 | Hurt |
| structured RAG baselines (GraphRAG, LightRAG) vs embedding RAG | GPT-4o QA · EpBench-2000 | F1: 0.675 -> 0.544 / 0.494 | Hurt |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- HippoRAG2 — HippoRAG: Neurobiologically Inspired Long-Term Memory for Large Language Models (NeurIPS 2024) · better on 6 of 6 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Episodic Memories Generation and Evaluation Benchmark for Large Language Models (ICLR 2025 · similarity 0.90)
- Human-inspired Episodic Memory for Infinite Context LLMs (ICLR 2025 · similarity 0.89)
- REMem: Reasoning with Episodic Memory in Language Agent (ICLR 2026 · similarity 0.88)
- E-mem: Multi-Agent Based Episodic Context Reconstruction for LLM Agent Memory (ICML 2026 · similarity 0.88)
- MC-Search: Evaluating and Enhancing Multimodal Agentic Search with Structured Long Reasoning Chains (ICLR 2026 · similarity 0.87)
- Beyond RAG for Agent Memory: Retrieval by Decoupling and Aggregation (NeurIPS 2026 · similarity 0.87)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
