Unverified50% confidenceFactExact time
After receiving a query, the query-driven planner determines what evidence is needed, whether the evidence is already in memory, and if not, which historical frames need to be revisited, outputting an evidence acquisition path rather than an answer.
1
Sources
50%
Confidence
Long-term
Relevance
10/15/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedFor systems that rely on polling external data sources on a schedule, conflating non-responses with negative answers pollutes downstream analysis; one should first check how many queries actually returned a result before examining the content of those results.73% similarUnverifiedWhen the system makes an error, responsibility decomposition should be performed first: determine whether the retrieved content contains the answer; if it does, the problem lies in the generation or context organization stage; if it does not, trace back to whether it is a data gap or a query understanding deviation.72% similarUnverifiedTo track and reproduce failures, each evaluation case needs to record the question, expected answer, retrieved document IDs and chunks, similarity scores, prompt and model configuration, generated answer, metric results, latency, and token usage, while avoiding recording keys or sensitive user data.71% similarUnverifiedThe decision layer backend adopts a Bayesian Beta posterior, updating based on historical results in memory, and confidence moves as evidence accumulates.71% similarUnverifiedEvidence recall is the proportion of the source text fragments on which the final answer depends that were actually read in the trajectory, and it can be used to distinguish retrieval failures from transcription failures.70% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1039499API
curl https://kongchang.com/api/v1/knowledge/claims/1039499MCP
get_claim(id=1039499)