paper
Trace Only What You Need: Structure-Aware On-Demand Hypergraph Memory for Long-Document Question Answering
paperactiveprovisional
trace-only-what-you-need-structure-aware-on-demand-hypergraph-memory-for-long-document-question-answering-c8d517e8·1 events·first seen 7d agoAliases: Trace Only What You Need: Structure-Aware On-Demand Hypergraph Memory for Long-Document Question Answering
Co-occurring entities
More like this (12)
Document Visual Question AnsweringUncertainty-Aware Hybrid Retrieval for Long-Document RAGMulti-hop Question AnsweringBias Benchmark for Question AnsweringAttention Amnesia in Hybrid LLMs: When CoT Fine-Tuning Breaks Long-Range Recall, and How to Fix ItHierarchical Graph MemoryA Training-Free Mixture-of-Agents Framework for Multi-Document Summarization using LLMs and Knowledge GraphsAttention Expansion: Enhancing Keyphrase Extraction from Long Documents with Attention-Augmented Contextualized Embeddingsgraph-based memory architectureWhen to Write and When to Suppress: Route-Specialized Dual Adapters for Memory-Assisted Knowledge EditingBeyond the Commitment Boundary: Probing Epiphenomenal Chain-of-Thought in Large Reasoning ModelsLong-context Reasoning Benchmarks
Recent events (1)
DocTrace: Structure-Aware On-Demand Hypergraph Memory for Long-Document QA
Researchers introduce DocTrace, a multi-agent RAG framework for long-document question answering that uses query-triggered knowledge organization rather than costly query-agnostic preprocessing. The system combines a lightweight document structural tree index, on-demand hypergraph working memory, and a graph-structured experience memory that stores successful reasoning plans for reuse. Evaluated on four long-document QA datasets, DocTrace outperforms the strongest baseline (ComoRAG) by up to 8.85% F1 and 4.40% EM while reducing computational cost by 53.32%.