InfMem: Learning System-2 Memory Control for Long-Context Agent
- Published
- Source
- arXiv
- Paper number
- 117
- Field
- Memory / Long Context
- arXiv ID
- 2602.02704
Key points
- Large language models struggle with effective multi-hop reasoning over ultra-long documents because of the middle loss phenomenon and the fidelity dilemma of compression.
- Existing memory-augmented agents often rely on passive and reactive memory update strategies, which risks losing decisive but low-salience evidence needed for complex reasoning.
- Ultra-long document processing introduces high compute costs and memory overhead, making current solutions inefficient for practical deployment.
- InfMem implements a control protocol of pre-think, retrieve, and record, allowing the agent to actively monitor evidence sufficiency, synthesize retrieval queries, and perform evidence-aware coupled compression.
- It enables non-monotonic access to any part of the document so that it can fetch missing evidence for the current query, including targeted retrieval inside the document.
- The framework uses an adaptive early-stopping mechanism that ends the memory update loop once sufficient evidence has accumulated, optimizing compute efficiency.
Paper links
External research summaries. These are not HDATF publications or measured product results.