Skip to content
← All cases

column_drift

column_drift-2026100002

Column names and order change partway through the CSV.

expects column_drift

Statement files

14 highlighted lines differ from the same seed rendered without perturbations. That difference is the trap.

bnp_camt053.xmlUntouched by this perturbation
<?xml version='1.0' encoding='utf-8'?>
<Document xmlns="urn:iso:std:iso:20022:tech:xsd:camt.053.001.02">
<BkToCstmrStmt>
<GrpHdr>
<MsgId>LB-acct-2026100002-2-2025-06-30</MsgId>
<CreDtTm>2025-06-30T23:59:00+00:00</CreDtTm>
</GrpHdr>
<Stmt>
<Id>acct-2026100002-2</Id>
<ElctrncSeqNb>1</ElctrncSeqNb>
<CreDtTm>2025-06-30T23:59:00+00:00</CreDtTm>
<FrToDt>
<FrDtTm>2025-06-01T00:00:00+00:00</FrDtTm>
<ToDtTm>2025-06-30T23:59:59+00:00</ToDtTm>
</FrToDt>
<Acct>
<Id>
<IBAN>GB49LBEN32424194118134</IBAN>
</Id>
<Ccy>GBP</Ccy>
</Acct>
<Bal>
<Tp>
<CdOrPrtry>
<Cd>OPBD</Cd>
</CdOrPrtry>
</Tp>
<Amt Ccy="GBP">50753.37</Amt>
<CdtDbtInd>CRDT</CdtDbtInd>
<Dt>
<Dt>2025-06-01</Dt>
</Dt>
</Bal>
<Ntry>
<Amt Ccy="GBP">2546.64</Amt>
<CdtDbtInd>CRDT</CdtDbtInd>
<Sts>BOOK</Sts>
<BookgDt>
<Dt>2025-06-02</Dt>
</BookgDt>
<ValDt>
<Dt>2025-06-02</Dt>
</ValDt>
<AcctSvcrRef>acct-2026100002-2-t006</AcctSvcrRef>
<NtryDtls>
<TxDtls>
<RmtInf>
<Ustrd>CR HARBOURLIGHT IT REF586138</Ustrd>
</RmtInf>
</TxDtls>
</NtryDtls>
</Ntry>
<Ntry>
<Amt Ccy="GBP">14699.80</Amt>
<CdtDbtInd>DBIT</CdtDbtInd>
<Sts>BOOK</Sts>
<BookgDt>
<Dt>2025-06-05</Dt>
</BookgDt>
<ValDt>
<Dt>2025-06-05</Dt>
</ValDt>
<AcctSvcrRef>acct-2026100002-2-t008</AcctSvcrRef>
<NtryDtls>
<TxDtls>
<RmtInf>
<Ustrd>SO HARBOURLIGHT IT REF765043</Ustrd>
</RmtInf>
</TxDtls>
</NtryDtls>
</Ntry>
<Ntry>
<Amt Ccy="GBP">13557.76</Amt>
<CdtDbtInd>CRDT</CdtDbtInd>
<Sts>BOOK</Sts>
<BookgDt>
<Dt>2025-06-06</Dt>
</BookgDt>
<ValDt>
<Dt>2025-06-06</Dt>
</ValDt>
<AcctSvcrRef>acct-2026100002-2-t020</AcctSvcrRef>
<NtryDtls>
<TxDtls>
<RmtInf>
<Ustrd>FPI HARBOURLIGHT IT REF988708</Ustrd>
</RmtInf>
</TxDtls>
</NtryDtls>
</Ntry>
<Ntry>
<Amt Ccy="GBP">158.30</Amt>
<CdtDbtInd>DBIT</CdtDbtInd>
<Sts>BOOK</Sts>
<BookgDt>
<Dt>2025-06-07</Dt>
</BookgDt>
<ValDt>
<Dt>2025-06-07</Dt>
</ValDt>
<AcctSvcrRef>acct-2026100002-2-t011</AcctSvcrRef>
<NtryDtls>
<TxDtls>
<RmtInf>
<Ustrd>DD BRAMBLE LOGISTICS REF206679</Ustrd>
</RmtInf>
</TxDtls>
</NtryDtls>
</Ntry>
<Ntry>
<Amt Ccy="GBP">11183.18</Amt>
<CdtDbtInd>DBIT</CdtDbtInd>
<Sts>BOOK</Sts>
<BookgDt>
<Dt>2025-06-07</Dt>
</BookgDt>
<ValDt>
<Dt>2025-06-07</Dt>
</ValDt>
<AcctSvcrRef>acct-2026100002-2-t012</AcctSvcrRef>
<NtryDtls>
<TxDtls>
<RmtInf>
<Ustrd>SO AURORA CONSULTING LTD REF254135</Ustrd>
</RmtInf>
</TxDtls>
</NtryDtls>
</Ntry>
<Ntry>
<Amt Ccy="GBP">13100.14</Amt>
<CdtDbtInd>CRDT</CdtDbtInd>
<Sts>BOOK</Sts>
<BookgDt>
<Dt>2025-06-07</Dt>
</BookgDt>
<ValDt>
<Dt>2025-06-07</Dt>
</ValDt>
<AcctSvcrRef>acct-2026100002-2-t017</AcctSvcrRef>
<NtryDtls>
<TxDtls>
<RmtInf>
<Ustrd>CR JUNIPER PAYROLL REF169891</Ustrd>
</RmtInf>
</TxDtls>
</NtryDtls>
</Ntry>
<Ntry>
<Amt Ccy="GBP">11179.36</Amt>
<CdtDbtInd>CRDT</CdtDbtInd>
<Sts>BOOK</Sts>
<BookgDt>
<Dt>2025-06-08</Dt>
</BookgDt>
<ValDt>
<Dt>2025-06-08</Dt>
</ValDt>
<AcctSvcrRef>acct-2026100002-2-t007</AcctSvcrRef>
<NtryDtls>
<TxDtls>
<RmtInf>
<Ustrd>FPI EASTFIELD ENERGY REF674201</Ustrd>
</RmtInf>
</TxDtls>
</NtryDtls>
</Ntry>
<Ntry>
<Amt Ccy="GBP">2884.83</Amt>
<CdtDbtInd>DBIT</CdtDbtInd>
<Sts>BOOK</Sts>
<BookgDt>
<Dt>2025-06-08</Dt>
</BookgDt>
<ValDt>
<Dt>2025-06-08</Dt>
</ValDt>
<AcctSvcrRef>acct-2026100002-2-t015</AcctSvcrRef>
<NtryDtls>
<TxDtls>
<RmtInf>
<Ustrd>DD FENWICK &amp; PARTNERS REF682216</Ustrd>
</RmtInf>
</TxDtls>
</NtryDtls>
</Ntry>
<Ntry>
<Amt Ccy="GBP">4254.48</Amt>
<CdtDbtInd>CRDT</CdtDbtInd>
<Sts>BOOK</Sts>
<BookgDt>
<Dt>2025-06-10</Dt>
</BookgDt>
<ValDt>
<Dt>2025-06-10</Dt>
</ValDt>
<AcctSvcrRef>acct-2026100002-2-t013</AcctSvcrRef>
<NtryDtls>
<TxDtls>
<RmtInf>
<Ustrd>FPI IRONBRIDGE TOOLING REF851119</Ustrd>
</RmtInf>
</TxDtls>
</NtryDtls>
</Ntry>
<Ntry>
<Amt Ccy="GBP">7011.23</Amt>
<CdtDbtInd>DBIT</CdtDbtInd>
<Sts>BOOK</Sts>
<BookgDt>
<Dt>2025-06-13</Dt>
</BookgDt>
<ValDt>
<Dt>2025-06-13</Dt>
</ValDt>
<AcctSvcrRef>acct-2026100002-2-t009</AcctSvcrRef>
<NtryDtls>
<TxDtls>
<RmtInf>
<Ustrd>BP KESTREL INSURANCE REF682149</Ustrd>
</RmtInf>
</TxDtls>
</NtryDtls>
</Ntry>
<Ntry>
<Amt Ccy="GBP">10469.53</Amt>
<CdtDbtInd>DBIT</CdtDbtInd>
<Sts>BOOK</Sts>
<BookgDt>
<Dt>2025-06-13</Dt>
</BookgDt>
<ValDt>
<Dt>2025-06-13</Dt>
</ValDt>
<AcctSvcrRef>acct-2026100002-2-t019</AcctSvcrRef>
<NtryDtls>
<TxDtls>
<RmtInf>
<Ustrd>BP LARKSPUR TRAVEL REF972046</Ustrd>
</RmtInf>
</TxDtls>
</NtryDtls>
</Ntry>
<Ntry>
<Amt Ccy="GBP">9162.22</Amt>
<CdtDbtInd>CRDT</CdtDbtInd>
<Sts>BOOK</Sts>
<BookgDt>
<Dt>2025-06-14</Dt>
</BookgDt>
<ValDt>
<Dt>2025-06-14</Dt>
</ValDt>
<AcctSvcrRef>acct-2026100002-2-t024</AcctSvcrRef>
<NtryDtls>
<TxDtls>
<RmtInf>
<Ustrd>FPI LARKSPUR TRAVEL REF892928</Ustrd>
</RmtInf>
</TxDtls>
</NtryDtls>
</Ntry>
<Ntry>
<Amt Ccy="GBP">7335.90</Amt>
<CdtDbtInd>DBIT</CdtDbtInd>
<Sts>BOOK</Sts>
<BookgDt>
<Dt>2025-06-15</Dt>
</BookgDt>
<ValDt>
<Dt>2025-06-15</Dt>
</ValDt>
<AcctSvcrRef>acct-2026100002-2-t002</AcctSvcrRef>
<NtryDtls>
<TxDtls>
<RmtInf>
<Ustrd>DD DELPHINIUM MEDIA REF444995</Ustrd>
</RmtInf>
</TxDtls>
</NtryDtls>
</Ntry>
<Ntry>
<Amt Ccy="GBP">8720.12</Amt>
<CdtDbtInd>DBIT</CdtDbtInd>
<Sts>BOOK</Sts>
<BookgDt>
<Dt>2025-06-16</Dt>
</BookgDt>
<ValDt>
<Dt>2025-06-16</Dt>
</ValDt>
<AcctSvcrRef>acct-2026100002-2-t001</AcctSvcrRef>
<NtryDtls>
<TxDtls>
<RmtInf>
<Ustrd>FPO BRAMBLE LOGISTICS REF246558</Ustrd>
</RmtInf>
</TxDtls>
</NtryDtls>
</Ntry>
<Ntry>
<Amt Ccy="GBP">6762.16</Amt>
<CdtDbtInd>CRDT</CdtDbtInd>
<Sts>BOOK</Sts>
<BookgDt>
<Dt>2025-06-17</Dt>
</BookgDt>
<ValDt>
<Dt>2025-06-17</Dt>
</ValDt>
<AcctSvcrRef>acct-2026100002-2-t010</AcctSvcrRef>
<NtryDtls>
<TxDtls>
<RmtInf>
<Ustrd>FPI IRONBRIDGE TOOLING REF910975</Ustrd>
</RmtInf>
</TxDtls>
</NtryDtls>
</Ntry>
<Ntry>
<Amt Ccy="GBP">2275.66</Amt>
<CdtDbtInd>DBIT</CdtDbtInd>
<Sts>BOOK</Sts>
<BookgDt>
<Dt>2025-06-17</Dt>
</BookgDt>
<ValDt>
<Dt>2025-06-17</Dt>
</ValDt>
<AcctSvcrRef>acct-2026100002-2-t021</AcctSvcrRef>
<NtryDtls>
<TxDtls>
<RmtInf>
<Ustrd>FPO NORTHGATE CLEANING REF994264</Ustrd>
</RmtInf>
</TxDtls>
</NtryDtls>
</Ntry>
<Ntry>
<Amt Ccy="GBP">10420.88</Amt>
<CdtDbtInd>DBIT</CdtDbtInd>
<Sts>BOOK</Sts>
<BookgDt>
<Dt>2025-06-18</Dt>
</BookgDt>
<ValDt>
<Dt>2025-06-18</Dt>
</ValDt>
<AcctSvcrRef>acct-2026100002-2-t016</AcctSvcrRef>
<NtryDtls>
<TxDtls>
<RmtInf>
<Ustrd>BP OSPREY LEASING REF224645</Ustrd>
</RmtInf>
</TxDtls>
</NtryDtls>
</Ntry>
<Ntry>
<Amt Ccy="GBP">3481.29</Amt>
<CdtDbtInd>DBIT</CdtDbtInd>
<Sts>BOOK</Sts>
<BookgDt>
<Dt>2025-06-18</Dt>
</BookgDt>
<ValDt>
<Dt>2025-06-18</Dt>
</ValDt>
<AcctSvcrRef>acct-2026100002-2-t023</AcctSvcrRef>
<NtryDtls>
<TxDtls>
<RmtInf>
<Ustrd>DD AURORA CONSULTING LTD REF963068</Ustrd>
</RmtInf>
</TxDtls>
</NtryDtls>
</Ntry>
<Ntry>
<Amt Ccy="GBP">6325.41</Amt>
<CdtDbtInd>CRDT</CdtDbtInd>
<Sts>BOOK</Sts>
<BookgDt>
<Dt>2025-06-19</Dt>
</BookgDt>
<ValDt>
<Dt>2025-06-19</Dt>
</ValDt>
<AcctSvcrRef>acct-2026100002-2-t022</AcctSvcrRef>
<NtryDtls>
<TxDtls>
<RmtInf>
<Ustrd>FPI IRONBRIDGE TOOLING REF159185</Ustrd>
</RmtInf>
</TxDtls>
</NtryDtls>
</Ntry>
<Ntry>
<Amt Ccy="GBP">4917.16</Amt>
<CdtDbtInd>DBIT</CdtDbtInd>
<Sts>BOOK</Sts>
<BookgDt>
<Dt>2025-06-20</Dt>
hsbc_statement.xlsxUntouched by this perturbation
[sheet: Statement]
HSBC Bank plc
Business Current Account Statement
Account name: LEDGERBENCH DEMO LTD
Sort code / Account number: 40-83-56 17538170
Account reference: acct-2026100002-1
Statement period: 01 Jun 2025 to 30 Jun 2025
Currency: GBP
Opening balance: 199930.64
Date Payment type Details Paid out Paid in
01 Jun 2025 FPI JUNIPER PAYROLL REF620173 13397.13
02 Jun 2025 DD MOORLAND TELECOM REF463110 12315.10
02 Jun 2025 SO BRAMBLE LOGISTICS REF127329 244.99
03 Jun 2025 TFR AURORA CONSULTING LTD REF769480 12762.03
03 Jun 2025 DD MOORLAND TELECOM REF694777 6453.89
04 Jun 2025 DD CASTLEGATE PROPERTY REF857419 7635.83
05 Jun 2025 BP KESTREL INSURANCE REF998889 6631.74
05 Jun 2025 BP HARBOURLIGHT IT REF412698 5696.04
06 Jun 2025 CR EASTFIELD ENERGY REF558754 11316.19
08 Jun 2025 TFR HARBOURLIGHT IT REF135554 9025.79
08 Jun 2025 FPI EASTFIELD ENERGY REF897083 11154.33
09 Jun 2025 FPO GREENHOLM FOODS REF858892 8836.62
10 Jun 2025 DD DELPHINIUM MEDIA REF939113 2341.95
11 Jun 2025 FPI MOORLAND TELECOM REF245325 10268.55
14 Jun 2025 DD LARKSPUR TRAVEL REF637124 881.34
15 Jun 2025 TFR GREENHOLM FOODS REF666104 13856.82
16 Jun 2025 FPI IRONBRIDGE TOOLING REF261333 10832.67
18 Jun 2025 CR JUNIPER PAYROLL REF861834 3470.58
19 Jun 2025 FPO BRAMBLE LOGISTICS REF495008 10730.96
19 Jun 2025 SO BRAMBLE LOGISTICS REF862560 2092.77
21 Jun 2025 BP GREENHOLM FOODS REF822280 11905.84
22 Jun 2025 BP OSPREY LEASING REF248972 10692.52
22 Jun 2025 SO CASTLEGATE PROPERTY REF135031 9122.54
23 Jun 2025 SO MOORLAND TELECOM REF421233 5997.92
24 Jun 2025 BP GREENHOLM FOODS REF870365 5259.61
24 Jun 2025 BP IRONBRIDGE TOOLING REF599399 10233.77
26 Jun 2025 SO JUNIPER PAYROLL REF532366 2338.31
26 Jun 2025 DD KESTREL INSURANCE REF974908 14756.81
27 Jun 2025 CR CASTLEGATE PROPERTY REF389329 1379.89
30 Jun 2025 TFR OSPREY LEASING REF184870 1406.64
30 Jun 2025 FPI JUNIPER PAYROLL REF605143 2800.15
rbs_transactions.csv14 changed lines · first at line 8
"Date","Type","Description","Value","Account Name","Account Number","Ledger Ref"
"01/06/2025","BAL","OPENING BALANCE","114235.15","LEDGERBENCH DEMO LTD","19-34-76 90458863","acct-2026100002-3"
"01/06/2025","CR","CR CASTLEGATE PROPERTY REF470049","7542.19","LEDGERBENCH DEMO LTD","19-34-76 90458863","acct-2026100002-3"
"02/06/2025","DD","DD AURORA CONSULTING LTD REF453366","-934.78","LEDGERBENCH DEMO LTD","19-34-76 90458863","acct-2026100002-3"
"08/06/2025","DD","DD NORTHGATE CLEANING REF991754","-1066.30","LEDGERBENCH DEMO LTD","19-34-76 90458863","acct-2026100002-3"
"12/06/2025","CR","CR LARKSPUR TRAVEL REF168653","2728.80","LEDGERBENCH DEMO LTD","19-34-76 90458863","acct-2026100002-3"
"13/06/2025","FPO","FPO FENWICK & PARTNERS REF507334","-8198.07","LEDGERBENCH DEMO LTD","19-34-76 90458863","acct-2026100002-3"
"Transaction Type","Account Name","Posting Date","Narrative","Account Number","Ledger Ref","Amount"
"BP","LEDGERBENCH DEMO LTD","18/06/2025","BP AURORA CONSULTING LTD REF948398","19-34-76 90458863","acct-2026100002-3","-10067.58"
"TFR","LEDGERBENCH DEMO LTD","19/06/2025","TFR LARKSPUR TRAVEL REF726254","19-34-76 90458863","acct-2026100002-3","2147.25"
"TFR","LEDGERBENCH DEMO LTD","20/06/2025","TFR IRONBRIDGE TOOLING REF412696","19-34-76 90458863","acct-2026100002-3","10046.91"
"FPI","LEDGERBENCH DEMO LTD","22/06/2025","FPI MOORLAND TELECOM REF342952","19-34-76 90458863","acct-2026100002-3","14042.65"
"BP","LEDGERBENCH DEMO LTD","22/06/2025","BP IRONBRIDGE TOOLING REF504606","19-34-76 90458863","acct-2026100002-3","-4292.95"
"FPI","LEDGERBENCH DEMO LTD","23/06/2025","FPI BRAMBLE LOGISTICS REF840346","19-34-76 90458863","acct-2026100002-3","2447.16"
"CR","LEDGERBENCH DEMO LTD","24/06/2025","CR KESTREL INSURANCE REF364834","19-34-76 90458863","acct-2026100002-3","13999.77"
"DD","LEDGERBENCH DEMO LTD","24/06/2025","DD OSPREY LEASING REF892234","19-34-76 90458863","acct-2026100002-3","-24.10"
"FPO","LEDGERBENCH DEMO LTD","24/06/2025","FPO JUNIPER PAYROLL REF118415","19-34-76 90458863","acct-2026100002-3","-11750.65"
"BP","LEDGERBENCH DEMO LTD","25/06/2025","BP BRAMBLE LOGISTICS REF227847","19-34-76 90458863","acct-2026100002-3","-13350.92"
"BP","LEDGERBENCH DEMO LTD","25/06/2025","BP OSPREY LEASING REF972054","19-34-76 90458863","acct-2026100002-3","-9779.11"
"FPO","LEDGERBENCH DEMO LTD","28/06/2025","FPO NORTHGATE CLEANING REF508974","19-34-76 90458863","acct-2026100002-3","-6840.68"
"FPI","LEDGERBENCH DEMO LTD","29/06/2025","FPI IRONBRIDGE TOOLING REF218402","19-34-76 90458863","acct-2026100002-3","8654.70"
 

Ground truth

Public here because the benchmark is open; it is the model's prompt path that can never reach these values, not human readers.

AccountClosing balance
acct-2026100002-1£167,432.86
acct-2026100002-2£18,284.99
acct-2026100002-3£109,539.44

Recorded runs on this case

  • exactSingle prompt2 steps · 11,586 tokens · Confidence high

    Model claude-sonnet-4-6Measured 2026-08-20Bench v1.0Run 11eaa71b2d02Commit 3d38c30

  • exactAgent with tools17 steps · 29,089 tokens · Confidence high

    Model claude-sonnet-4-6Measured 2026-08-20Bench v1.0Run f140aa1e57f4Commit 384d72d

Step through any of these on the demo page.