Unverified50% confidenceOpinionExact time
End-to-end prompting methods entangle the three components of evidence retrieval, reasoning, and uncertainty estimation, making it difficult to pinpoint the source of errors.
1
Sources
50%
Confidence
Long-term
Relevance
9/14/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedPassing brief summaries is low-cost but may erase critical evidence and quietly turn inferences into facts, preventing downstream agents from determining whether information represents an explicit decision or a model guess.71% similarUnverifiedInference methods based on model internal features face a trade-off between false positives and false negatives70% similarUnverifiedThe core idea of hard negative mining is to introduce samples near the classification boundary to drive refinement of the decision boundary, enabling the model to learn deep semantic features rather than surface-level keyword patterns.69% similarUnverifiedDiagnostic counterexample feedback provides concrete evidence that directly locates problem boundaries, making it more effective than generic prompts or simply informing the agent that an error exists.69% similarUnverifiedWhen a certain delta happens to be empty tool-call data, if the parsing logic handles it improperly, it may incorrectly judge that the entire reasoning stream has ended and thus lose subsequent content.68% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/919708API
curl https://kongchang.com/api/v1/knowledge/claims/919708MCP
get_claim(id=919708)