Unverified50% confidenceFactExact time
Loreflow has a stricter sub-budget for live/federated evidence (the highest-risk third-party text): at most 3 of the 8 slots.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
10/16/2026
First Seen
Valid until: 1/14/2027
Sources
Related Entities
Related Claims
UnverifiedEvery Loreflow answer is assembled under a fixed evidence budget: at most 8 pieces of evidence, each 1000 code points, totaling 8000 code points, counted by code point rather than UTF-16 units.72% similarUnverifiedThe main limitations of the two-stage prompt include: longer output, subjective risk judgments, and coverage only of what the model actually read (web scraping may return a rewritten version rather than the original text).60% similarUnverifiedIn large-context-window scenarios, the risk of prompt injection is amplified, because the model must read a large amount of external material at once, and malicious instructions may be scattered in different positions within the long text and be harder to find via keyword filtering.59% similarUnverifiedThe systematic review field widely recognizes that the cost of missing critical literature (false negatives) far exceeds that of incorrectly including irrelevant literature (false positives), and therefore the evaluation framework must place recall at its core.59% similarUnverifiedThere are three reasons for not directly sending all documents to a large model: the model cannot read all content due to context window limitations, inference costs are high, and inference speed is slow.59% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1041419API
curl https://kongchang.com/api/v1/knowledge/claims/1041419MCP
get_claim(id=1041419)