Dr-DCI: Scaling Direct Corpus Interaction via Dynamic Workspace Expansion
Quick Answer
DR-DCI enhances Direct Corpus Interaction by dynamically expanding local workspaces for efficient document retrieval and operations, achieving 71.2% accuracy on Browsecomp-Plus and outperforming BM25 in large corpora.
Quick Take
DR-DCI enhances Direct Corpus Interaction by dynamically expanding local workspaces for efficient document retrieval and operations, achieving 71.2% accuracy on Browsecomp-Plus and outperforming BM25 in large corpora. Its design allows for scalable exploration while maintaining precision, with accuracy improving to 73.3% through context resets.
Key Points
- DR-DCI achieves 71.2% accuracy, improving by up to 8.3 points over raw DCI.
- Performance remains stable from 100K to 10M documents, unlike raw DCI.
- Ablation analysis highlights the importance of ranked previews and inter-document DCI.
- Context reset feature boosts accuracy to 73.3% on Browsecomp-Plus.
- DR-DCI scores 63.0 on Wiki-18 QA across six benchmarks.
Paper Resources
Article Content
From source RSS / original summaryarXiv:2606. 14885v1 Announce Type: new Abstract: Agentic search over large corpora relies on retriever-mediated interfaces (e. g. , BM25 or ColBERT) for scalable candidate discovery. While effective at ranking relevant documents, these interfaces expose evidence only as ranked results or bounded document views, limiting agents' ability to reorganize material and verify constraints across documents.
Direct Corpus Interaction (DCI) addresses this limitation by exposing shell-executable corpus operations for flexible search, filtering, comparison, and verification. However, full-corpus terminal commands become slow and unstable as the corpus grows, degrading performance and efficiency. We introduce DR-DCI, a retriever-steered DCI framework that treats retrieval as an agent-callable action for expanding a local workspace.
Rather than operating directly over the full corpus, the agent dynamically pulls relevant documents into an evolving workspace and conducts DCI operations within it. This design combines retriever-level recall with DCI-style precision: retrieval keeps exploration scalable, while DCI preserves the local operations needed for effective evidence resolution. Experiments show that DR-DCI is both effective and efficient across scales. On Browsecomp-Plus, DR-DCI reaches 71.
2\% accuracy, improving over raw DCI and ablated variants by up to 8. 3 points while reducing tool usage, wall time, and estimated cost. With workspace-preserving context reset, accuracy further improves to 73. 3\%. In corpus-scaling experiments, DR-DCI remains effective from 100K to 10M documents, whereas raw DCI becomes unstable and BM25 performs substantially worse. DR-DCI also scales to a 20M-scale file-per-document Wiki-18 QA setting, achieving an average score of 63.
0 across six benchmarks and outperforming retrieval-based and trained search-agent baselines. Ablation analysis further shows that ranked previews and inter-document DCI are key to performance.
Reader Mode unavailable (could not extract clean content).
Want this in your inbox every morning?
Daily brief at your local 8am — bilingual EN/中文, free.
More from arXiv cs.AI
See more →Arbor: Tree Search as a Cognition Layer for Autonomous Agents
Arbor introduces a multi-agent framework utilizing structured tree search for optimizing LLM inference, achieving up to 193% throughput-latency improvement compared to vendor-optimized systems. It employs an Orchestrator and Critic agent for stability and coordination, demonstrating hardware-agnostic performance with minimal variance.