Unverified50% confidenceBenchmarkExact time
In the experiment, the task_score under the shaped reward configuration was only 0.125, compared to 0.417 for the naive reward, with an effect size of d=4.47.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
9/7/2026
First Seen
Valid until: 12/6/2026
Sources
Related Entities
Related Claims
UnverifiedThe Mann-Whitney U test shows that task_score under the naive survival reward falls below the random-policy baseline, with p=0.012.73% similarUnverifiedThe strict task success rate for novice users is 15%, and 28%–33% for intermediate and above users; when encountering difficulties, the ultimate success rate for novices is only 4%, compared to 15% for experts63% similarUnverifiedThe power-law exponent between test loss and compute is approximately 0.05, meaning that every 10x increase in compute yields approximately a 12% improvement in model performance.63% similarUnverifiedEven GPT-4-level Agents often achieve end-to-end success rates below 60% on complex tasks involving more than 5 tool calls.63% similarUnverified斯坦福大学的AgentBench基准测试（2023年）量化证明，在复杂多步骤任务中纯ReAct范式的成功率往往不超过30%62% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/870228API
curl https://kongchang.com/api/v1/knowledge/claims/870228MCP
get_claim(id=870228)