Unverified50% confidenceBenchmarkExact time
On the ID/OOD average success rate on ALFWorld, BranchOPD improves over Vanilla OPD by 13.8 percentage points.
1
Sources
50%
Confidence
Long-term
Relevance
10/12/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedAt 64× compression, AdaMem's average relative gain expands to 14.6%, with a maximum improvement of 9.8 percentage points (19.7%) on the PopQA dataset.61% similarUnverifiedBranchOPD outperforms both Vanilla OPD and the more expensive Multi-Rollout OPD (n=7) on all evaluation benchmarks.59% similarUnverifiedWhen OSWorld was released in 2024, the strongest baseline model score was only approximately 12%, while the human completion rate was approximately 72%.57% similarUnverifiedRoboDojo具身智能测评显示，最强模型得分仅12.8分（人类基准为100分），仿真成功率不足9%，真机成功率12.8%54% similarUnverifiedAnthropic的Computer Use在OSWorld基准测试上的操作准确率约为14.9%（2024年10月数据）51% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1001144API
curl https://kongchang.com/api/v1/knowledge/claims/1001144MCP
get_claim(id=1001144)