Unverified50% confidenceFactExact time
At the analytically derived break-even probability of 0.80, seven prompting configurations chose to abstain on all tickets, while Haiku few-shot routed only 19/1,000.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
10/23/2026
First Seen
Valid until: 1/21/2027
Sources
Related Entities
Related Claims
UnverifiedRanked by question difficulty, persistence rate was 91.4% on the 60 questions that all models answered correctly, but only 73.5% on the 52 questions that at least one model got wrong, indicating that sycophancy tracks uncertainty.69% similarUnverifiedIn an S&P 500 option chain experiment, after sparsifying the data to retain only 20 quotes and introducing no-arbitrage constraints, the out-of-sample extrapolation error on the held-out set decreased by 22% and the calendar arbitrage occurrence rate dropped from 10% to 1%.68% similarUnverifiedThe more detailed the answers provided to the AI during the requirements analysis phase, the lower the probability of bugs in the final product; the cost of fixing errors found in the requirements phase is only 1/10 to 1/100 of the cost of finding the same type of error in the coding phase.68% similarUnverifiedThe leaderboard score is the no-arbitrage rate, i.e., the proportion of families in which the bookmaker can lock in at most 5 cents; models with coverage below 90% are not scored.67% similarUnverifiedIn a standalone test of the "view incomplete" prompt on 100 questions, the wrong-answer rate was 16% without the prompt and dropped to 5% with it.66% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1088398API
curl https://kongchang.com/api/v1/knowledge/claims/1088398MCP
get_claim(id=1088398)