Unverified50% confidenceFactExact time
Probabilistic judgments (such as whether an answer is harmful, or whether it cites a document that does not exist) have no single correct answer and must rely on model semantic understanding, with cost and latency several orders of magnitude higher.
1
Sources
50%
Confidence
Long-term
Relevance
10/13/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedThe study observed a clear lack of correlation between the correctness of answers and the validity of reasoning traces, with models frequently arriving at correct answers even when the reasoning traces are completely invalid.76% similarUnverifiedCorrectly answering intervention-type questions requires additionally introducing domain knowledge, causal graphs, or carefully designed experiments (such as randomized controlled trials); data and computational power alone are insufficient.76% similarUnverifiedThe model's tool-triggering decision is essentially a probabilistic threshold problem: when confidence exceeds the threshold the model tends to answer directly, and only triggers retrieval when confidence falls below it.76% similarUnverifiedBecause the output of large models is non-deterministic, simple assert statements cannot determine whether generated text is correct, which is why evaluations (evals) are needed.76% similarUnverifiedThe output of large models is probabilistic and cannot be simply judged for equality, so they cannot be tested using traditional assertion methods.76% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1019247API
curl https://kongchang.com/api/v1/knowledge/claims/1019247MCP
get_claim(id=1019247)