Research map / Large language models / Retrieval and knowledge
Failure is Feedback: History-Aware Backtracking for Agentic Traversal in Multimodal Graphs
Joohyung Yun, Doyup Lee, Wook-Shin Han
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 |
|---|---|---|---|
| add history-aware backtracking orchestration (ablation) | FiF · MultimodalQA (10% subset) | Recall@10: 79.26 -> 85.82 | Improved |
| add history-aware backtracking orchestration (ablation) | FiF · WebQA (10% subset) | Recall@10: 77.45 -> 81.91 | Improved |
| add LLM reasoning in traversal agent (ablation) | FiF · MultimodalQA (10% subset) | Recall@10: 78.04 -> 85.82 | Improved |
| add LLM reasoning in traversal agent (ablation) | FiF · WebQA (10% subset) | Recall@10: 78.92 -> 81.91 | Improved |
| add global hop (ablation) | FiF · MultimodalQA (10% subset) | Recall@10: 75.79 -> 85.82 | Improved |
| add global hop (ablation) | FiF · WebQA (10% subset) | Recall@10: 73.30 -> 81.91 | Improved |
| add adaptive vector-search granularity (ablation) | FiF · MultimodalQA (10% subset) | Recall@10: 83.16 -> 85.82 | Improved |
| add adaptive vector-search granularity (ablation) | FiF · WebQA (10% subset) | Recall@10: 79.21 -> 81.91 | Improved |
| add subquery planner (replanning) (ablation) | FiF · MultimodalQA (10% subset) | Recall@10: 76.2 -> 85.82 | Improved |
| add subquery planner (replanning) (ablation) | FiF · WebQA (10% subset) | Recall@10: 76.38 -> 81.91 | Improved |
| add history-aware backtracking orchestration (efficiency) (ablation) | FiF · MultimodalQA (10% subset) | Time (ms): 211,951 -> 117,651 | Improved |
| add subquery planner replanning (cost) (ablation) | FiF · MultimodalQA (10% subset) | # LLM calls: 5.31 -> 7.97 | 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.
- NV-Embed-v2 — NV-Embed: Improved Techniques for Training LLMs as Generalist Embedding Models (ICLR 2025) · better on 3 of 3 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
- Navigating Massive Visual Context in Retrieval-Augmented Generation via Multimodal Memory Graph (ICML 2026 · similarity 0.87)
- Resolving Evidence Sparsity: Agentic Context Engineering for Long-Document Understanding (CVPR 2026 · similarity 0.87)
- ARK: A Dual-Axis Multimodal Retrieval Benchmark along Reasoning and Knowledge (NeurIPS 2026 · similarity 0.86)
- TopoGraphRAG-Bench: Evaluating Multimodal GraphRAG on Layout-Grounded Evidence Reasoning (NeurIPS 2026 · similarity 0.86)
- MMSearch-Plus: Benchmarking Provenance-Aware Search for Multimodal Browsing Agents (ICLR 2026 · similarity 0.85)
- Beyond RAG for Agent Memory: Retrieval by Decoupling and Aggregation (NeurIPS 2026 · similarity 0.85)
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.
