Unverified50% confidenceFactExact time
The paper proposes a retrieval-based automatic labeling strategy: for each extracted fact, if it is retrieved and included in the reply in subsequent interactions it is labeled as a positive example, otherwise as a negative example.
1
Sources
50%
Confidence
Long-term
Relevance
10/19/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedMulti-hop retrieval means first retrieving a first batch of documents, constructing a new query based on the results and then retrieving the next batch, gradually assembling the complete evidence chain; it applies to the situation where the all-evidence score is low while any-hit is normal.74% similarUnverifiedGenerative retrieval allows models to directly generate document identifiers or answers, while agentic retrieval endows systems with the ability to autonomously plan, conduct multi-round retrieval, and evaluate result quality.73% similarVerifiedIf the results of automated checking tools are linked to acceptance decisions, a robust appeals and manual review mechanism must be established.72% similarUnverifiedEvery interaction between a user and an AI model forms a retrievable query record, and these records can serve as direct evidence in compliance investigations.72% similarUnverifiedWhen dealing with highly time-sensitive information, models with internet search or reasoning capabilities should be prioritized.70% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1065519API
curl https://kongchang.com/api/v1/knowledge/claims/1065519MCP
get_claim(id=1065519)