Unverified50% confidenceFactExact time
The Sieve and Sage framework decouples retrieval failure into two states: Unanswerable and Distracted.
1
Sources
50%
Confidence
Long-term
Relevance
9/30/2026
First Seen
Sources
Related Entities
Related Claims
Unverified语义层与执行层的割裂是业务Skill失效的根本原因，语义层描述模糊会导致Agent从不调用或错误场景频繁触发该Skill65% similarUnverifiedUsing incorrect default capability advertisement values masks the true situation and leads to difficult-to-reproduce errors, whereas unknown is an explicit semantic state that callers can use for graceful degradation64% similarVerifiedCurrent multi-agent frameworks face challenges such as the accumulation of information transfer errors between agents, limited ability to handle complex business logic, and the need for human oversight to verify the reliability of generated results.61% similarUnverifiedLoaded questions contain unverified false assumptions, and LLMs find it difficult to identify such traps due to a lack of metacognitive ability.61% similarUnverifiedTraditional NLI tasks assume a pair of sentences is already given and only require judging whether they are entailment, neutral, or contradiction, whereas the core problem in multi-document contradiction discovery is finding which pairs of sentences constitute candidates.60% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/961110API
curl https://kongchang.com/api/v1/knowledge/claims/961110MCP
get_claim(id=961110)