Unverified50% confidenceOpinionExact time
In multi-hop reasoning tasks, the success rate of retrieval-type tasks (single-hop) is far higher than that of reasoning-type tasks (multi-hop).
1
Sources
50%
Confidence
Long-term
Relevance
10/14/2026
First Seen
Sources
Related Entities
Related Claims
VerifiedIncreasing the number of inference passes incurs higher computational costs but in exchange lowers the complexity of each individual task—this is the trade-off of the grouped extraction strategy.76% similarUnverified提升模型单步精度的边际收益远不如缩短任务链条、增加检查点74% similarVerifiedUnder the same model, the gap in task success rate between a professional prompt and a vague prompt can be several times larger74% similarUnverifiedAgentic retrieval involves multiple rounds of model calls and multiple knowledge base queries, resulting in higher token consumption and latency compared to single retrieval.72% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1028484API
curl https://kongchang.com/api/v1/knowledge/claims/1028484MCP
get_claim(id=1028484)