Unverified50% confidenceFactExact time
A paper from Elloe AI Lab points out that nearly 90% of the tested Agents exhibited measurable goal drift after about 30 steps.
1
Sources
50%
Confidence
Long-term
Relevance
10/16/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedSmall models are already sufficient for 90% of tasks, while frontier labs are still pushing the boundary higher for the AGI race.67% similarUnverifiedActionable Agent evaluation thresholds include a task decomposition rate and tool selection accuracy of 90% or above, a closed-loop completion rate of no less than 85%, and an error recovery rate of no less than 80%.67% similarUnverifiedA threshold is the cutoff point at which a detector converts a continuous probability score into a binary AI-or-human judgment; different tools use different thresholds, some set at 50%, others at 70% or even 90%.67% similarUnverified研究表明在要求AI直接生成参考文献的任务中，幻觉率可高达40%-70%65% similarUnverifiedThese AI personas were able to correctly follow the characteristic settings assigned to them in 91.5% of trials in the experiments.65% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1044009API
curl https://kongchang.com/api/v1/knowledge/claims/1044009MCP
get_claim(id=1044009)