ResearchBenchmarksStructured video reasoning on Sliding Puzzle OverallFollow86.5Exact Match (%)TBS-2037.6250.316375.69Jun 11, 2026Evaluation ResultsMethodMethodLinksExact Match (%)Normalized Time Cost (×)TBS-20Budget=20Budget=202026.0686.55.67BoN-20Budget=20Budget=202026.0682.26.35TBS-10Budget=10Budget=102026.0680.84.11BoN-10Budget=10Budget=102026.0674.74.27BoN-5Budget=5Budget=52026.0667.52.81BoN-1Budget=1Budget=12026.0639.51