Research map / Learning theory and algorithms / Combinatorial and search optimization
Depth-First Proof-Number Search with Heuristic Edge Cost and Application to Chemical Synthesis Planning
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 heuristic edge cost initialization (ablation) | DFPN · Retrosynthesis, 1945 patent targets (897 hard instances) | total runtime (s): 20,133 -> 5,654 | Helped |
| add heuristic edge cost initialization (ablation) | DFPN · Retrosynthesis, 1945 patent targets (897 hard instances) | instances solved: 770 -> 842 | Helped |
| use heuristic edge costs instead of heuristic leaf initialization (ablation) | DFPNH · Retrosynthesis, 1945 patent targets (897 hard instances) | total runtime (s): 30,822 -> 5,654 | Helped |
| heuristic leaf initialization of proof numbers (ablation) | DFPN · Retrosynthesis, 1945 patent targets (897 hard instances) | total runtime (s): 20,133 -> 30,822 | Hurt |
| heuristic edge costs instead of unit edge cost (ablation) | DFPN-1 · Retrosynthesis, 1945 patent targets (897 hard instances) | instances solved: 691 -> 842 | Helped |
| heuristic edge costs instead of unit edge cost (ablation) | DFPN-1 · Retrosynthesis, 1945 patent targets (897 hard instances) | total runtime (s): 46,542 -> 5,654 | Helped |
| DFPN-E instead of enhanced MCTS | MCTS · Retrosynthesis, 1945 patent targets (897 hard instances) | total runtime (s): 18,552 -> 5,654 | Helped |
| DFPN-E instead of enhanced MCTS | MCTS · Retrosynthesis, 1945 patent targets (897 hard instances) | instances solved: 852 -> 842 | Hurt |
| heuristic edge costs (ablation) | DFPN · Retrosynthesis, 1945 patent targets (897 hard instances) | average pathway length: 227.42 -> 5.72 | Helped |
Closest papers by meaning
- SeeA*: Efficient Exploration-Enhanced A* Search by Selective Sampling (NeurIPS 2024 · similarity 0.87)
- Double-Ended Synthesis Planning with Goal-Constrained Bidirectional Search (NeurIPS 2024 · similarity 0.86)
- MOTIF: Multi-strategy Optimization via Turn-based Interactive Framework (AAAI 2026 · similarity 0.85)
- Improving Monte Carlo Tree Search for Symbolic Regression (NeurIPS 2025 · similarity 0.84)
- LLM-Augmented Chemical Synthesis and Design Decision Programs (ICML 2025 · similarity 0.84)
- Retro-fallback: retrosynthetic planning in an uncertain world (ICLR 2024 · similarity 0.83)
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.