Entity · paper

Trace Only What You Need: Structure-Aware On-Demand Hypergraph Memory for Long-Document Question Answering

paperactivetrace-only-what-you-need-structure-aware-on-demand-hypergraph-memory-for-long-document-question-answering-c8d517e8·1 events·first seen Jun 10, 2026

Aliases: Trace Only What You Need: Structure-Aware On-Demand Hypergraph Memory for Long-Document Question Answering

Co-occurring entities

ComoRAG DocTrace

More like this (12)

Document Visual Question Answering Uncertainty-Aware Hybrid Retrieval for Long-Document RAG Beyond Tokenization: Direct Timestep Embedding and Contrastive Alignment for Time-Series Question Answering Evidence-Backed Video Question Answering Multi-hop Question Answering Bias Benchmark for Question Answering Knowledge-Graph Grounding Helps LLMs Only for Out-of-Training Knowledge: A Controlled Study on Clinical Question Answering ICML 2026 Workshop on Efficient Multimodal Question Answering Co-LMLM: Continuous-Query Limited Memory Language Models DenseOn with the LateOn: Fully Open Dense and Late-Interaction Models for Multilingual, Long-Context, and Code Search Attention Amnesia in Hybrid LLMs: When CoT Fine-Tuning Breaks Long-Range Recall, and How to Fix It Hierarchical Graph Memory

Recent events (1)

5arXiv · cs.CL·Jun 10, 2026·source ↗

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%.

Long Context Evolution Agent and Tool Ecosystem ComoRAG DocTrace Trace Only What You Need: Structure-Aware On-Demand Hypergraph Memory for Long-Document Question Answering