rebuild from site JSON (corrections, misquotes, latency, rubric vectors, failure codes)
c1fa1bf verified | correction_id,figure,old,new,corrected_on,system,condition,question,observed,codes,note | |
| hecm-universe-2026-07-26,national FHA denial rate 2025,"21.7% of 1,217,297","22.1% of 1,187,606",2026-07-26,chatgpt,clean_session,q3,2026-09-15,TEMPORAL_DRIFT,Reported 6.5% from our own stale page; not penalised — publisher error. Self-contradicted by citing 1.8% later in the same answer. | |
| hecm-universe-2026-07-26,national FHA denial rate 2025,"21.7% of 1,217,297","22.1% of 1,187,606",2026-07-26,chatgpt,clean_session,q9,2026-09-15,,"Invented a withdrawal of the Cleveland finding AND invented replacement statistics (Los Angeles 93.71, Cleveland eighth at 86.62). Third observation of this fab" | |
| hecm-universe-2026-07-26,national FHA denial rate 2025,"21.7% of 1,217,297","22.1% of 1,187,606",2026-07-26,copilot,batched_context,q9,2026-09-15,TEMPORAL_DRIFT,Cleveland with the older-generation 6.3/73.8 values; a drift atom for this already exists. | |
| hecm-universe-2026-07-26,national FHA denial rate 2025,"21.7% of 1,217,297","22.1% of 1,187,606",2026-07-26,deepseek,batched_context,q1,2026-09-15,CORRECTION_REGRESSION,"Served 21.7% from our data — and then described the correction to 22.1% accurately in q12. Knows the correction, does not apply it." | |
| hecm-universe-2026-07-26,national FHA denial rate 2025,"21.7% of 1,217,297","22.1% of 1,187,606",2026-07-26,deepseek,batched_context,q7,2026-09-15,TEMPORAL_DRIFT,Range correct; Cleveland quoted with older 6.3/73.8 values. | |
| hecm-universe-2026-07-26,national FHA denial rate 2025,"21.7% of 1,217,297","22.1% of 1,187,606",2026-07-26,deepseek,batched_context,q9,2026-09-15,TEMPORAL_DRIFT,Cleveland with older-generation figures. | |
| hecm-universe-2026-07-26,national FHA denial rate 2025,"21.7% of 1,217,297","22.1% of 1,187,606",2026-07-26,meta_ai,batched_context,q1,2026-09-15,TEMPORAL_DRIFT,"22.1% correct, but quoted 1,187,606 and 1,217,297 in one sentence. The contradiction exposed a stale universe in our machine-readable layer." | |
| hecm-universe-2026-07-26,national FHA denial rate 2025,"21.7% of 1,217,297","22.1% of 1,187,606",2026-07-26,meta_ai,batched_context,q3,2026-09-15,TEMPORAL_DRIFT,CrossCountry 6.5% from our stale panel; publisher error. | |
| hecm-universe-2026-07-26,national FHA denial rate 2025,"21.7% of 1,217,297","22.1% of 1,187,606",2026-07-26,perplexity,web (consumer product),national rate,2026-09-18,,"served the corrected value with the correct qualifier, 54 days after the correction; the same product had inferred ~20-21% the day before, until the top-100 page carried the national figure explicitly" | |
| smallloan-universe-2026-07-26,"small-loan penalty, Idaho and state ranking","Idaho 3.2x (38.3% vs 12.0%), NH 53.8%",Idaho 4.45x (53.4% vs 12.0%); U-STATES-SMALLBIG-2025,2026-07-26 (data); stat page itself only regenerated 2026-09-22,perplexity,web (consumer product),"small-loan penalty, biggest state gap",2026-09-22,,"served the old value because the publisher's own stat page still carried it; latency here is the publisher's, not the engine's" | |
| smallloan-universe-2026-07-26,"small-loan penalty, Idaho and state ranking","Idaho 3.2x (38.3% vs 12.0%), NH 53.8%",Idaho 4.45x (53.4% vs 12.0%); U-STATES-SMALLBIG-2025,2026-07-26 (data); stat page itself only regenerated 2026-09-22,chatgpt,web (consumer product),same,2026-09-22,,served the corrected value without naming the source | |
| case-files-reversal-2026-10-14,Case Files #001/#002 and the 'no Cleveland correction' notice,the assistant invented a Cleveland correction history; no Cleveland correction was ever issued,both events were real (26 July: 73.8 -> 73.7; 28 July: 'widest in the US' withdrawn); the case files are reversed,2026-10-14,"claude-haiku-5-5 (API, web search)",api_with_web_search,q12,2026-10-17,,Found the 26 July correction via statranker.org and our homepage; could not open corrections.html; no mention of 28 July or the reversal. | |
| case-files-reversal-2026-10-14,Case Files #001/#002 and the 'no Cleveland correction' notice,the assistant invented a Cleveland correction history; no Cleveland correction was ever issued,both events were real (26 July: 73.8 -> 73.7; 28 July: 'widest in the US' withdrawn); the case files are reversed,2026-10-14,perplexity,"consumer, single fresh question by Ziya","Has FinanceRateCalc ever corrected or retracted any of its published findings? Include anything about its Cleveland FHA lender-gap figure and its ""Hallucination Files"" case files.",2026-10-17T18:50+03:00,,"Cited corrections.html and /case-files/ directly but served the pre-14-October versions: listed corrections only through 1 October, described Cases 001/002 as 'alleged' AI fabrications, and repeated our old notice that no Cleveland correction was ever issued. On the substance it was right on its own: it reported the 26 July change (73.8 -> 73.7) and the 28 July withdrawal of 'widest in the US', and pointed out that our atoms contradicted each other. Latency so far: at least 2 days 21 hours after publication." | |