{"data":{"kind":"file","path":"README.md","version_id":"u533f0opg2ji2uwz3vqvj1t0","entry":{"name":"README.md","path":"README.md","is_directory":false,"size":1058,"modified_at":"2026-08-08T14:31:37.177000","content_hash":"2db641856386dfe2b68bc0ed4cfb0c0a7fba917325510495d5916bd63334107c"},"entries":[],"content":"# Graph Navigation & Path Finding\n\nA verifiers environment that evaluates LLM reasoning on graph traversal and shortest path problems.\n\n## Task Types\n\n1. **Dijkstra Shortest Path** — Find the minimum-cost path in a weighted undirected graph\n2. **BFS Shortest Hop** — Find the minimum-hop path in an unweighted graph\n3. **Grid Navigation** — Navigate from start (S) to end (E) on a 2D grid avoiding obstacles\n4. **Multi-Hop Routing** — Visit all waypoints in optimal order with minimum total cost\n\n## Scoring\n\n| Metric | Weight | Description |\n|--------|--------|-------------|\n| Path Correctness | 0.60 | Whether the found path matches or is near-optimal |\n| Explanation Quality | 0.25 | Reasoning depth and algorithm understanding |\n| Format Compliance | 0.15 | Adherence to requested answer format |\n\n## Usage\n\n```python\nimport verifiers as vf\nenv = vf.load_environment(\"malisa/graph-navigation-pathfinding\")\n```\n\n## Parameters\n\n- `num_examples`: Number of problems to generate (default: 60)\n- `seed`: Random seed for reproducibility (default: 42)\n","encoding":"utf-8","truncated":false,"total_bytes":1058},"status":null}