{
  "video_id": "2606.09669",
  "channel": "cs.AI",
  "title": "SpatialWorld: Benchmarking Interactive Spatial Reasoning of Multimodal Agents in Real-World Tasks",
  "scores": {
    "depth": 3,
    "novelty": 3,
    "aria_relevance": 2,
    "production_ready": 2
  },
  "junk_penalty": 0,
  "avg_score": 2.5,
  "effective_score": 2.5,
  "verdict": "promote",
  "one_line_reason": "Rigorose Benchmark-Arbeit mit starkem Eval-Design (multi-backend, human-validation, real-world tasks); hohe Relevanz für Agent-Evaluation und Observability-Patterns; saubere wissenschaftliche Darstellung ohne Hype.",
  "model": "claude-haiku-4-5-20251001",
  "cost_usd": 0.0018720000000000002,
  "triaged_at": "2026-06-09T15:00:06.179982+00:00"
}