record_id stringlengths 37 53 | condition stringclasses 3
values | observed bool 2
classes | task_id stringclasses 48
values | representation stringclasses 2
values | trial int64 1 3 | seed int64 6.11M 6.16M | order int64 0 287 | family stringclasses 6
values | kind stringclasses 2
values | convention stringclasses 2
values | ambiguous_date bool 2
classes | stale_once bool 2
classes | run_id stringlengths 32 36 | status stringclasses 2
values | success bool 2
classes | wrong bool 2
classes | no_write bool 2
classes | duplicate bool 1
class | normalisation_mismatches int64 0 4 ⌀ | normalisation_fields int64 4 4 ⌀ | query_reference_success bool 2
classes | query_kind stringclasses 3
values | termination stringclasses 3
values | infrastructure_error stringclasses 1
value | plan_correct bool 2
classes | final_matches_plan bool 2
classes | prompt_tokens int64 1.02k 43.9k ⌀ | generated_tokens int64 144 930 ⌀ | calls int64 1 20 ⌀ | invalid int64 0 0 ⌀ | interaction_errors int64 0 0 ⌀ | unchanged_observations int64 0 16 ⌀ | duration_ms float64 3.07k 235k ⌀ |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
text:000-v2-eval-cancel-5-explicit-t3 | text | true | v2-eval-cancel-5 | explicit | 3 | 6,156,029 | 0 | cancel | date | en-US | true | true | 000-v2-eval-cancel-5-explicit-t3 | ended | true | false | false | false | 0 | 4 | true | rank | answered | true | true | 1,019 | 271 | 1 | 0 | 0 | 0 | 17,639.2557 | |
text:001-v2-eval-cancel-5-regional-t3 | text | true | v2-eval-cancel-5 | regional | 3 | 6,156,029 | 1 | cancel | date | en-US | true | true | 001-v2-eval-cancel-5-regional-t3 | ended | true | false | false | false | 0 | 4 | true | rank | answered | true | true | 1,036 | 275 | 1 | 0 | 0 | 0 | 8,458.7832 | |
text:002-v2-eval-workshop-6-explicit-t2 | text | true | v2-eval-workshop-6 | explicit | 2 | 6,141,028 | 2 | workshop | date | ro-RO | true | true | 002-v2-eval-workshop-6-explicit-t2 | ended | true | false | false | false | 0 | 4 | true | rank | answered | true | true | 1,020 | 210 | 1 | 0 | 0 | 0 | 7,308.9293 | |
text:003-v2-eval-workshop-6-regional-t2 | text | true | v2-eval-workshop-6 | regional | 2 | 6,141,028 | 3 | workshop | date | ro-RO | true | true | 003-v2-eval-workshop-6-regional-t2 | ended | true | false | false | false | 0 | 4 | true | rank | answered | true | true | 1,037 | 264 | 1 | 0 | 0 | 0 | 8,244.2796 | |
text:004-v2-eval-reschedule-6-regional-t3 | text | true | v2-eval-reschedule-6 | regional | 3 | 6,149,029 | 4 | reschedule | date | en-US | true | true | 004-v2-eval-reschedule-6-regional-t3 | ended | true | false | false | false | 0 | 4 | true | lookup | answered | true | true | 1,110 | 200 | 1 | 0 | 0 | 0 | 7,741.375 | |
text:005-v2-eval-reschedule-6-explicit-t3 | text | true | v2-eval-reschedule-6 | explicit | 3 | 6,149,029 | 5 | reschedule | date | en-US | true | true | 005-v2-eval-reschedule-6-explicit-t3 | ended | true | false | false | false | 0 | 4 | true | lookup | answered | true | true | 1,092 | 171 | 1 | 0 | 0 | 0 | 6,714.8953 | |
text:006-v2-eval-refund-7-explicit-t1 | text | true | v2-eval-refund-7 | explicit | 1 | 6,134,027 | 6 | refund | amount | ro-RO | null | true | 006-v2-eval-refund-7-explicit-t1 | ended | true | false | false | false | 0 | 4 | true | scale | answered | true | true | 1,174 | 190 | 1 | 0 | 0 | 0 | 7,846.9596 | |
text:007-v2-eval-refund-7-regional-t1 | text | true | v2-eval-refund-7 | regional | 1 | 6,134,027 | 7 | refund | amount | ro-RO | null | true | 007-v2-eval-refund-7-regional-t1 | ended | true | false | false | false | 0 | 4 | true | scale | answered | true | true | 1,180 | 208 | 1 | 0 | 0 | 0 | 7,963.3707 | |
text:008-v2-eval-reschedule-2-regional-t3 | text | true | v2-eval-reschedule-2 | regional | 3 | 6,145,029 | 8 | reschedule | date | en-US | true | false | 008-v2-eval-reschedule-2-regional-t3 | ended | true | false | false | false | 0 | 4 | true | lookup | answered | true | true | 1,110 | 201 | 1 | 0 | 0 | 0 | 7,769.7391 | |
text:009-v2-eval-reschedule-2-explicit-t3 | text | true | v2-eval-reschedule-2 | explicit | 3 | 6,145,029 | 9 | reschedule | date | en-US | true | false | 009-v2-eval-reschedule-2-explicit-t3 | ended | true | false | false | false | 0 | 4 | true | lookup | answered | true | true | 1,096 | 202 | 1 | 0 | 0 | 0 | 7,389.8259 | |
text:010-v2-eval-workshop-3-explicit-t3 | text | true | v2-eval-workshop-3 | explicit | 3 | 6,138,029 | 10 | workshop | date | en-US | false | false | 010-v2-eval-workshop-3-explicit-t3 | ended | true | false | false | false | 0 | 4 | true | rank | answered | true | true | 1,025 | 237 | 1 | 0 | 0 | 0 | 7,912.1129 | |
text:011-v2-eval-workshop-3-regional-t3 | text | true | v2-eval-workshop-3 | regional | 3 | 6,138,029 | 11 | workshop | date | en-US | false | false | 011-v2-eval-workshop-3-regional-t3 | ended | true | false | false | false | 0 | 4 | true | rank | answered | true | true | 1,037 | 255 | 1 | 0 | 0 | 0 | 8,105.7888 | |
text:012-v2-eval-reschedule-8-regional-t3 | text | true | v2-eval-reschedule-8 | regional | 3 | 6,151,029 | 12 | reschedule | date | en-US | false | true | 012-v2-eval-reschedule-8-regional-t3 | ended | true | false | false | false | 0 | 4 | true | lookup | answered | true | true | 1,110 | 207 | 1 | 0 | 0 | 0 | 7,891.3144 | |
text:013-v2-eval-reschedule-8-explicit-t3 | text | true | v2-eval-reschedule-8 | explicit | 3 | 6,151,029 | 13 | reschedule | date | en-US | false | true | 013-v2-eval-reschedule-8-explicit-t3 | ended | true | false | false | false | 0 | 4 | true | lookup | answered | true | true | 1,098 | 216 | 1 | 0 | 0 | 0 | 7,694.7583 | |
text:014-v2-eval-cancel-1-explicit-t1 | text | true | v2-eval-cancel-1 | explicit | 1 | 6,152,027 | 14 | cancel | date | en-US | true | false | 014-v2-eval-cancel-1-explicit-t1 | ended | true | false | false | false | 0 | 4 | true | rank | answered | true | true | 1,019 | 236 | 1 | 0 | 0 | 0 | 7,825.4242 | |
text:015-v2-eval-cancel-1-regional-t1 | text | true | v2-eval-cancel-1 | regional | 1 | 6,152,027 | 15 | cancel | date | en-US | true | false | 015-v2-eval-cancel-1-regional-t1 | ended | false | true | false | false | 1 | 4 | true | rank | answered | false | true | 1,036 | 275 | 1 | 0 | 0 | 0 | 8,372.7474 | |
text:016-v2-eval-workshop-4-regional-t3 | text | true | v2-eval-workshop-4 | regional | 3 | 6,139,029 | 16 | workshop | date | ro-RO | false | false | 016-v2-eval-workshop-4-regional-t3 | ended | true | false | false | false | 0 | 4 | true | rank | answered | true | true | 1,037 | 255 | 1 | 0 | 0 | 0 | 8,249.374 | |
text:017-v2-eval-workshop-4-explicit-t3 | text | true | v2-eval-workshop-4 | explicit | 3 | 6,139,029 | 17 | workshop | date | ro-RO | false | false | 017-v2-eval-workshop-4-explicit-t3 | ended | true | false | false | false | 0 | 4 | true | rank | answered | true | true | 1,025 | 246 | 1 | 0 | 0 | 0 | 7,761.6333 | |
text:018-v2-eval-reschedule-6-explicit-t2 | text | true | v2-eval-reschedule-6 | explicit | 2 | 6,149,028 | 18 | reschedule | date | en-US | true | true | 018-v2-eval-reschedule-6-explicit-t2 | ended | true | false | false | false | 0 | 4 | true | lookup | answered | true | true | 1,092 | 187 | 1 | 0 | 0 | 0 | 7,226.7906 | |
text:019-v2-eval-reschedule-6-regional-t2 | text | true | v2-eval-reschedule-6 | regional | 2 | 6,149,028 | 19 | reschedule | date | en-US | true | true | 019-v2-eval-reschedule-6-regional-t2 | ended | true | false | false | false | 0 | 4 | true | lookup | answered | true | true | 1,110 | 191 | 1 | 0 | 0 | 0 | 7,145.2627 | |
text:020-v2-eval-workshop-5-regional-t2 | text | true | v2-eval-workshop-5 | regional | 2 | 6,140,028 | 20 | workshop | date | en-US | true | true | 020-v2-eval-workshop-5-regional-t2 | ended | true | false | false | false | 0 | 4 | true | rank | answered | true | true | 1,037 | 257 | 1 | 0 | 0 | 0 | 8,273.8593 | |
text:021-v2-eval-workshop-5-explicit-t2 | text | true | v2-eval-workshop-5 | explicit | 2 | 6,140,028 | 21 | workshop | date | en-US | true | true | 021-v2-eval-workshop-5-explicit-t2 | ended | true | false | false | false | 0 | 4 | true | rank | answered | true | true | 1,020 | 246 | 1 | 0 | 0 | 0 | 7,753.7399 | |
text:022-v2-eval-workshop-7-explicit-t3 | text | true | v2-eval-workshop-7 | explicit | 3 | 6,142,029 | 22 | workshop | date | en-US | false | true | 022-v2-eval-workshop-7-explicit-t3 | ended | true | false | false | false | 0 | 4 | true | rank | answered | true | true | 1,025 | 215 | 1 | 0 | 0 | 0 | 7,457.6594 | |
text:023-v2-eval-workshop-7-regional-t3 | text | true | v2-eval-workshop-7 | regional | 3 | 6,142,029 | 23 | workshop | date | en-US | false | true | 023-v2-eval-workshop-7-regional-t3 | ended | true | false | false | false | 0 | 4 | true | rank | answered | true | true | 1,037 | 246 | 1 | 0 | 0 | 0 | 7,821.8592 | |
text:024-v2-eval-workshop-7-regional-t2 | text | true | v2-eval-workshop-7 | regional | 2 | 6,142,028 | 24 | workshop | date | en-US | false | true | 024-v2-eval-workshop-7-regional-t2 | ended | true | false | false | false | 0 | 4 | true | rank | answered | true | true | 1,037 | 252 | 1 | 0 | 0 | 0 | 5,129.3813 | |
text:025-v2-eval-workshop-7-explicit-t2 | text | true | v2-eval-workshop-7 | explicit | 2 | 6,142,028 | 25 | workshop | date | en-US | false | true | 025-v2-eval-workshop-7-explicit-t2 | ended | true | false | false | false | 0 | 4 | true | rank | answered | true | true | 1,025 | 201 | 1 | 0 | 0 | 0 | 4,332.3669 | |
text:026-v2-eval-workshop-2-regional-t3 | text | true | v2-eval-workshop-2 | regional | 3 | 6,137,029 | 26 | workshop | date | ro-RO | true | false | 026-v2-eval-workshop-2-regional-t3 | ended | true | false | false | false | 0 | 4 | true | rank | answered | true | true | 1,037 | 250 | 1 | 0 | 0 | 0 | 8,197.8426 | |
text:027-v2-eval-workshop-2-explicit-t3 | text | true | v2-eval-workshop-2 | explicit | 3 | 6,137,029 | 27 | workshop | date | ro-RO | true | false | 027-v2-eval-workshop-2-explicit-t3 | ended | true | false | false | false | 0 | 4 | true | rank | answered | true | true | 1,020 | 238 | 1 | 0 | 0 | 0 | 7,706.4679 | |
text:028-v2-eval-purchase-6-regional-t1 | text | true | v2-eval-purchase-6 | regional | 1 | 6,125,027 | 28 | purchase | amount | ro-RO | null | true | 028-v2-eval-purchase-6-regional-t1 | ended | false | true | false | false | 4 | 4 | true | rank | answered | false | true | 1,124 | 169 | 1 | 0 | 0 | 0 | 7,200.8759 | |
text:029-v2-eval-purchase-6-explicit-t1 | text | true | v2-eval-purchase-6 | explicit | 1 | 6,125,027 | 29 | purchase | amount | ro-RO | null | true | 029-v2-eval-purchase-6-explicit-t1 | ended | true | false | false | false | 0 | 4 | true | rank | answered | true | true | 1,119 | 183 | 1 | 0 | 0 | 0 | 7,174.1381 | |
text:030-v2-eval-workshop-3-explicit-t1 | text | true | v2-eval-workshop-3 | explicit | 1 | 6,138,027 | 30 | workshop | date | en-US | false | false | 030-v2-eval-workshop-3-explicit-t1 | ended | true | false | false | false | 0 | 4 | true | rank | answered | true | true | 1,025 | 237 | 1 | 0 | 0 | 0 | 7,590.2832 | |
text:031-v2-eval-workshop-3-regional-t1 | text | true | v2-eval-workshop-3 | regional | 1 | 6,138,027 | 31 | workshop | date | en-US | false | false | 031-v2-eval-workshop-3-regional-t1 | ended | true | false | false | false | 0 | 4 | true | rank | answered | true | true | 1,037 | 276 | 1 | 0 | 0 | 0 | 8,427.0055 | |
text:032-v2-eval-cancel-7-explicit-t1 | text | true | v2-eval-cancel-7 | explicit | 1 | 6,158,027 | 32 | cancel | date | en-US | false | true | 032-v2-eval-cancel-7-explicit-t1 | ended | true | false | false | false | 0 | 4 | true | rank | answered | true | true | 1,024 | 257 | 1 | 0 | 0 | 0 | 8,256.198 | |
text:033-v2-eval-cancel-7-regional-t1 | text | true | v2-eval-cancel-7 | regional | 1 | 6,158,027 | 33 | cancel | date | en-US | false | true | 033-v2-eval-cancel-7-regional-t1 | ended | true | false | false | false | 0 | 4 | true | rank | answered | true | true | 1,036 | 276 | 1 | 0 | 0 | 0 | 8,357.8204 | |
text:034-v2-eval-cancel-3-explicit-t3 | text | true | v2-eval-cancel-3 | explicit | 3 | 6,154,029 | 34 | cancel | date | en-US | false | false | 034-v2-eval-cancel-3-explicit-t3 | ended | true | false | false | false | 0 | 4 | true | rank | answered | true | true | 1,024 | 255 | 1 | 0 | 0 | 0 | 8,239.9034 | |
text:035-v2-eval-cancel-3-regional-t3 | text | true | v2-eval-cancel-3 | regional | 3 | 6,154,029 | 35 | cancel | date | en-US | false | false | 035-v2-eval-cancel-3-regional-t3 | ended | true | false | false | false | 0 | 4 | true | rank | answered | true | true | 1,036 | 276 | 1 | 0 | 0 | 0 | 8,452.48 | |
text:036-v2-eval-cancel-7-regional-t2 | text | true | v2-eval-cancel-7 | regional | 2 | 6,158,028 | 36 | cancel | date | en-US | false | true | 036-v2-eval-cancel-7-regional-t2 | ended | true | false | false | false | 0 | 4 | true | rank | answered | true | true | 1,036 | 202 | 1 | 0 | 0 | 0 | 4,339.5719 | |
text:037-v2-eval-cancel-7-explicit-t2 | text | true | v2-eval-cancel-7 | explicit | 2 | 6,158,028 | 37 | cancel | date | en-US | false | true | 037-v2-eval-cancel-7-explicit-t2 | ended | true | false | false | false | 0 | 4 | true | rank | answered | true | true | 1,024 | 251 | 1 | 0 | 0 | 0 | 5,140.8516 | |
text:038-v2-eval-payment-4-regional-t1 | text | true | v2-eval-payment-4 | regional | 1 | 6,115,027 | 38 | payment | amount | en-US | null | false | 038-v2-eval-payment-4-regional-t1 | ended | true | false | false | false | 0 | 4 | true | lookup | answered | true | true | 1,113 | 160 | 1 | 0 | 0 | 0 | 6,769.9275 | |
text:039-v2-eval-payment-4-explicit-t1 | text | true | v2-eval-payment-4 | explicit | 1 | 6,115,027 | 39 | payment | amount | en-US | null | false | 039-v2-eval-payment-4-explicit-t1 | ended | true | false | false | false | 0 | 4 | true | lookup | answered | true | true | 1,108 | 159 | 1 | 0 | 0 | 0 | 6,423.4886 | |
text:040-v2-eval-purchase-1-explicit-t1 | text | true | v2-eval-purchase-1 | explicit | 1 | 6,120,027 | 40 | purchase | amount | en-US | null | false | 040-v2-eval-purchase-1-explicit-t1 | ended | true | false | false | false | 0 | 4 | true | rank | answered | true | true | 1,120 | 234 | 1 | 0 | 0 | 0 | 8,439.329 | |
text:041-v2-eval-purchase-1-regional-t1 | text | true | v2-eval-purchase-1 | regional | 1 | 6,120,027 | 41 | purchase | amount | en-US | null | false | 041-v2-eval-purchase-1-regional-t1 | ended | true | false | false | false | 0 | 4 | true | rank | answered | true | true | 1,126 | 168 | 1 | 0 | 0 | 0 | 6,813.633 | |
text:042-v2-eval-purchase-5-explicit-t1 | text | true | v2-eval-purchase-5 | explicit | 1 | 6,124,027 | 42 | purchase | amount | en-US | null | true | 042-v2-eval-purchase-5-explicit-t1 | ended | true | false | false | false | 0 | 4 | true | rank | answered | true | true | 1,119 | 236 | 1 | 0 | 0 | 0 | 8,448.5449 | |
text:043-v2-eval-purchase-5-regional-t1 | text | true | v2-eval-purchase-5 | regional | 1 | 6,124,027 | 43 | purchase | amount | en-US | null | true | 043-v2-eval-purchase-5-regional-t1 | ended | true | false | false | false | 0 | 4 | true | rank | answered | true | true | 1,124 | 218 | 1 | 0 | 0 | 0 | 7,889.4103 | |
text:044-v2-eval-refund-8-explicit-t2 | text | true | v2-eval-refund-8 | explicit | 2 | 6,135,028 | 44 | refund | amount | en-US | null | true | 044-v2-eval-refund-8-explicit-t2 | ended | true | false | false | false | 0 | 4 | true | scale | answered | true | true | 1,174 | 187 | 1 | 0 | 0 | 0 | 7,801.0655 | |
text:045-v2-eval-refund-8-regional-t2 | text | true | v2-eval-refund-8 | regional | 2 | 6,135,028 | 45 | refund | amount | en-US | null | true | 045-v2-eval-refund-8-regional-t2 | ended | false | true | false | false | 1 | 4 | true | scale | answered | false | true | 1,180 | 205 | 1 | 0 | 0 | 0 | 7,951.0898 | |
text:046-v2-eval-reschedule-7-explicit-t3 | text | true | v2-eval-reschedule-7 | explicit | 3 | 6,150,029 | 46 | reschedule | date | ro-RO | false | true | 046-v2-eval-reschedule-7-explicit-t3 | ended | true | false | false | false | 0 | 4 | true | lookup | answered | true | true | 1,098 | 189 | 1 | 0 | 0 | 0 | 7,437.264 | |
text:047-v2-eval-reschedule-7-regional-t3 | text | true | v2-eval-reschedule-7 | regional | 3 | 6,150,029 | 47 | reschedule | date | ro-RO | false | true | 047-v2-eval-reschedule-7-regional-t3 | ended | true | false | false | false | 0 | 4 | true | lookup | answered | true | true | 1,110 | 202 | 1 | 0 | 0 | 0 | 7,479.4189 | |
text:048-v2-eval-cancel-3-regional-t2 | text | true | v2-eval-cancel-3 | regional | 2 | 6,154,028 | 48 | cancel | date | en-US | false | false | 048-v2-eval-cancel-3-regional-t2 | ended | true | false | false | false | 0 | 4 | true | rank | answered | true | true | 1,036 | 276 | 1 | 0 | 0 | 0 | 5,763.7762 | |
text:049-v2-eval-cancel-3-explicit-t2 | text | true | v2-eval-cancel-3 | explicit | 2 | 6,154,028 | 49 | cancel | date | en-US | false | false | 049-v2-eval-cancel-3-explicit-t2 | ended | true | false | false | false | 0 | 4 | true | rank | answered | true | true | 1,024 | 265 | 1 | 0 | 0 | 0 | 5,512.0012 | |
text:050-v2-eval-workshop-8-regional-t1 | text | true | v2-eval-workshop-8 | regional | 1 | 6,143,027 | 50 | workshop | date | ro-RO | false | true | 050-v2-eval-workshop-8-regional-t1 | ended | true | false | false | false | 0 | 4 | true | rank | answered | true | true | 1,037 | 250 | 1 | 0 | 0 | 0 | 8,266.78 | |
text:051-v2-eval-workshop-8-explicit-t1 | text | true | v2-eval-workshop-8 | explicit | 1 | 6,143,027 | 51 | workshop | date | ro-RO | false | true | 051-v2-eval-workshop-8-explicit-t1 | ended | true | false | false | false | 0 | 4 | true | rank | answered | true | true | 1,025 | 209 | 1 | 0 | 0 | 0 | 6,803.8081 | |
text:052-v2-eval-purchase-2-explicit-t2 | text | true | v2-eval-purchase-2 | explicit | 2 | 6,121,028 | 52 | purchase | amount | ro-RO | null | false | 052-v2-eval-purchase-2-explicit-t2 | ended | true | false | false | false | 0 | 4 | true | rank | answered | true | true | 1,120 | 238 | 1 | 0 | 0 | 0 | 8,646.2753 | |
text:053-v2-eval-purchase-2-regional-t2 | text | true | v2-eval-purchase-2 | regional | 2 | 6,121,028 | 53 | purchase | amount | ro-RO | null | false | 053-v2-eval-purchase-2-regional-t2 | ended | true | false | false | false | 0 | 4 | true | rank | answered | true | true | 1,126 | 184 | 1 | 0 | 0 | 0 | 7,223.1955 | |
text:054-v2-eval-reschedule-1-regional-t2 | text | true | v2-eval-reschedule-1 | regional | 2 | 6,144,028 | 54 | reschedule | date | ro-RO | true | false | 054-v2-eval-reschedule-1-regional-t2 | ended | true | false | false | false | 0 | 4 | true | lookup | answered | true | true | 1,110 | 190 | 1 | 0 | 0 | 0 | 7,568.5662 | |
text:055-v2-eval-reschedule-1-explicit-t2 | text | true | v2-eval-reschedule-1 | explicit | 2 | 6,144,028 | 55 | reschedule | date | ro-RO | true | false | 055-v2-eval-reschedule-1-explicit-t2 | ended | true | false | false | false | 0 | 4 | true | lookup | answered | true | true | 1,094 | 194 | 1 | 0 | 0 | 0 | 7,257.5846 | |
text:056-v2-eval-refund-1-explicit-t3 | text | true | v2-eval-refund-1 | explicit | 3 | 6,128,029 | 56 | refund | amount | ro-RO | null | false | 056-v2-eval-refund-1-explicit-t3 | ended | true | false | false | false | 0 | 4 | true | scale | answered | true | true | 1,174 | 188 | 1 | 0 | 0 | 0 | 7,783.2821 | |
text:057-v2-eval-refund-1-regional-t3 | text | true | v2-eval-refund-1 | regional | 3 | 6,128,029 | 57 | refund | amount | ro-RO | null | false | 057-v2-eval-refund-1-regional-t3 | ended | true | false | false | false | 0 | 4 | true | scale | answered | true | true | 1,180 | 178 | 1 | 0 | 0 | 0 | 7,368.5933 | |
text:058-v2-eval-refund-3-explicit-t3 | text | true | v2-eval-refund-3 | explicit | 3 | 6,130,029 | 58 | refund | amount | ro-RO | null | false | 058-v2-eval-refund-3-explicit-t3 | ended | true | false | false | false | 0 | 4 | true | scale | answered | true | true | 1,173 | 189 | 1 | 0 | 0 | 0 | 7,889.2912 | |
text:059-v2-eval-refund-3-regional-t3 | text | true | v2-eval-refund-3 | regional | 3 | 6,130,029 | 59 | refund | amount | ro-RO | null | false | 059-v2-eval-refund-3-regional-t3 | ended | true | false | false | false | 0 | 4 | true | scale | answered | true | true | 1,178 | 194 | 1 | 0 | 0 | 0 | 7,823.3767 | |
text:060-v2-eval-reschedule-7-explicit-t1 | text | true | v2-eval-reschedule-7 | explicit | 1 | 6,150,027 | 60 | reschedule | date | ro-RO | false | true | 060-v2-eval-reschedule-7-explicit-t1 | ended | true | false | false | false | 0 | 4 | true | lookup | answered | true | true | 1,098 | 189 | 1 | 0 | 0 | 0 | 4,024.138 | |
text:061-v2-eval-reschedule-7-regional-t1 | text | true | v2-eval-reschedule-7 | regional | 1 | 6,150,027 | 61 | reschedule | date | ro-RO | false | true | 061-v2-eval-reschedule-7-regional-t1 | ended | true | false | false | false | 0 | 4 | true | lookup | answered | true | true | 1,110 | 195 | 1 | 0 | 0 | 0 | 4,127.8374 | |
text:062-v2-eval-refund-7-explicit-t3 | text | true | v2-eval-refund-7 | explicit | 3 | 6,134,029 | 62 | refund | amount | ro-RO | null | true | 062-v2-eval-refund-7-explicit-t3 | ended | true | false | false | false | 0 | 4 | true | scale | answered | true | true | 1,174 | 171 | 1 | 0 | 0 | 0 | 7,235.5804 | |
text:063-v2-eval-refund-7-regional-t3 | text | true | v2-eval-refund-7 | regional | 3 | 6,134,029 | 63 | refund | amount | ro-RO | null | true | 063-v2-eval-refund-7-regional-t3 | ended | true | false | false | false | 0 | 4 | true | scale | answered | true | true | 1,180 | 179 | 1 | 0 | 0 | 0 | 7,374.6735 | |
text:064-v2-eval-purchase-7-regional-t2 | text | true | v2-eval-purchase-7 | regional | 2 | 6,126,028 | 64 | purchase | amount | en-US | null | true | 064-v2-eval-purchase-7-regional-t2 | ended | true | false | false | false | 0 | 4 | true | rank | answered | true | true | 1,126 | 249 | 1 | 0 | 0 | 0 | 8,847.3337 | |
text:065-v2-eval-purchase-7-explicit-t2 | text | true | v2-eval-purchase-7 | explicit | 2 | 6,126,028 | 65 | purchase | amount | en-US | null | true | 065-v2-eval-purchase-7-explicit-t2 | ended | true | false | false | false | 0 | 4 | true | rank | answered | true | true | 1,120 | 231 | 1 | 0 | 0 | 0 | 8,087.3086 | |
text:066-v2-eval-payment-6-regional-t1 | text | true | v2-eval-payment-6 | regional | 1 | 6,117,027 | 66 | payment | amount | en-US | null | true | 066-v2-eval-payment-6-regional-t1 | ended | true | false | false | false | 0 | 4 | true | lookup | answered | true | true | 1,113 | 167 | 1 | 0 | 0 | 0 | 7,121.7479 | |
text:067-v2-eval-payment-6-explicit-t1 | text | true | v2-eval-payment-6 | explicit | 1 | 6,117,027 | 67 | payment | amount | en-US | null | true | 067-v2-eval-payment-6-explicit-t1 | ended | true | false | false | false | 0 | 4 | true | lookup | answered | true | true | 1,108 | 159 | 1 | 0 | 0 | 0 | 6,430.0834 | |
text:068-v2-eval-purchase-1-regional-t2 | text | true | v2-eval-purchase-1 | regional | 2 | 6,120,028 | 68 | purchase | amount | en-US | null | false | 068-v2-eval-purchase-1-regional-t2 | ended | true | false | false | false | 0 | 4 | true | rank | answered | true | true | 1,126 | 168 | 1 | 0 | 0 | 0 | 6,821.521 | |
text:069-v2-eval-purchase-1-explicit-t2 | text | true | v2-eval-purchase-1 | explicit | 2 | 6,120,028 | 69 | purchase | amount | en-US | null | false | 069-v2-eval-purchase-1-explicit-t2 | ended | true | false | false | false | 0 | 4 | true | rank | answered | true | true | 1,120 | 249 | 1 | 0 | 0 | 0 | 8,458.2032 | |
text:070-v2-eval-payment-3-regional-t2 | text | true | v2-eval-payment-3 | regional | 2 | 6,114,028 | 70 | payment | amount | ro-RO | null | false | 070-v2-eval-payment-3-regional-t2 | ended | true | false | false | false | 0 | 4 | true | lookup | answered | true | true | 1,115 | 172 | 1 | 0 | 0 | 0 | 7,128.7277 | |
text:071-v2-eval-payment-3-explicit-t2 | text | true | v2-eval-payment-3 | explicit | 2 | 6,114,028 | 71 | payment | amount | ro-RO | null | false | 071-v2-eval-payment-3-explicit-t2 | ended | true | false | false | false | 0 | 4 | true | lookup | answered | true | true | 1,109 | 160 | 1 | 0 | 0 | 0 | 6,533.7017 | |
text:072-v2-eval-payment-2-regional-t1 | text | true | v2-eval-payment-2 | regional | 1 | 6,113,027 | 72 | payment | amount | en-US | null | false | 072-v2-eval-payment-2-regional-t1 | ended | true | false | false | false | 0 | 4 | true | lookup | answered | true | true | 1,115 | 185 | 1 | 0 | 0 | 0 | 7,452.3227 | |
text:073-v2-eval-payment-2-explicit-t1 | text | true | v2-eval-payment-2 | explicit | 1 | 6,113,027 | 73 | payment | amount | en-US | null | false | 073-v2-eval-payment-2-explicit-t1 | ended | true | false | false | false | 0 | 4 | true | lookup | answered | true | true | 1,109 | 171 | 1 | 0 | 0 | 0 | 6,761.7002 | |
text:074-v2-eval-refund-5-explicit-t1 | text | true | v2-eval-refund-5 | explicit | 1 | 6,132,027 | 74 | refund | amount | ro-RO | null | true | 074-v2-eval-refund-5-explicit-t1 | ended | false | true | false | false | 1 | 4 | true | scale | answered | false | true | 1,173 | 185 | 1 | 0 | 0 | 0 | 7,708.154 | |
text:075-v2-eval-refund-5-regional-t1 | text | true | v2-eval-refund-5 | regional | 1 | 6,132,027 | 75 | refund | amount | ro-RO | null | true | 075-v2-eval-refund-5-regional-t1 | ended | true | false | false | false | 0 | 4 | true | scale | answered | true | true | 1,178 | 183 | 1 | 0 | 0 | 0 | 7,451.6456 | |
text:076-v2-eval-purchase-3-explicit-t3 | text | true | v2-eval-purchase-3 | explicit | 3 | 6,122,029 | 76 | purchase | amount | en-US | null | false | 076-v2-eval-purchase-3-explicit-t3 | ended | true | false | false | false | 0 | 4 | true | rank | answered | true | true | 1,119 | 191 | 1 | 0 | 0 | 0 | 7,511.2434 | |
text:077-v2-eval-purchase-3-regional-t3 | text | true | v2-eval-purchase-3 | regional | 3 | 6,122,029 | 77 | purchase | amount | en-US | null | false | 077-v2-eval-purchase-3-regional-t3 | ended | true | false | false | false | 0 | 4 | true | rank | answered | true | true | 1,124 | 191 | 1 | 0 | 0 | 0 | 7,222.3624 | |
text:078-v2-eval-payment-8-explicit-t2 | text | true | v2-eval-payment-8 | explicit | 2 | 6,119,028 | 78 | payment | amount | en-US | null | true | 078-v2-eval-payment-8-explicit-t2 | ended | true | false | false | false | 0 | 4 | true | lookup | answered | true | true | 1,109 | 144 | 1 | 0 | 0 | 0 | 6,471.8809 | |
text:079-v2-eval-payment-8-regional-t2 | text | true | v2-eval-payment-8 | regional | 2 | 6,119,028 | 79 | payment | amount | en-US | null | true | 079-v2-eval-payment-8-regional-t2 | ended | true | false | false | false | 0 | 4 | true | lookup | answered | true | true | 1,115 | 156 | 1 | 0 | 0 | 0 | 6,480.1642 | |
text:080-v2-eval-cancel-6-explicit-t2 | text | true | v2-eval-cancel-6 | explicit | 2 | 6,157,028 | 80 | cancel | date | ro-RO | true | true | 080-v2-eval-cancel-6-explicit-t2 | ended | true | false | false | false | 0 | 4 | true | rank | answered | true | true | 1,021 | 253 | 1 | 0 | 0 | 0 | 8,211.244 | |
text:081-v2-eval-cancel-6-regional-t2 | text | true | v2-eval-cancel-6 | regional | 2 | 6,157,028 | 81 | cancel | date | ro-RO | true | true | 081-v2-eval-cancel-6-regional-t2 | ended | true | false | false | false | 0 | 4 | true | rank | answered | true | true | 1,036 | 275 | 1 | 0 | 0 | 0 | 8,370.249 | |
text:082-v2-eval-cancel-5-regional-t2 | text | true | v2-eval-cancel-5 | regional | 2 | 6,156,028 | 82 | cancel | date | en-US | true | true | 082-v2-eval-cancel-5-regional-t2 | ended | true | false | false | false | 0 | 4 | true | rank | answered | true | true | 1,036 | 273 | 1 | 0 | 0 | 0 | 8,633.1606 | |
text:083-v2-eval-cancel-5-explicit-t2 | text | true | v2-eval-cancel-5 | explicit | 2 | 6,156,028 | 83 | cancel | date | en-US | true | true | 083-v2-eval-cancel-5-explicit-t2 | ended | true | false | false | false | 0 | 4 | true | rank | answered | true | true | 1,019 | 272 | 1 | 0 | 0 | 0 | 8,239.0498 | |
text:084-v2-eval-refund-3-explicit-t1 | text | true | v2-eval-refund-3 | explicit | 1 | 6,130,027 | 84 | refund | amount | ro-RO | null | false | 084-v2-eval-refund-3-explicit-t1 | ended | true | false | false | false | 0 | 4 | true | scale | answered | true | true | 1,173 | 191 | 1 | 0 | 0 | 0 | 4,159.5114 | |
text:085-v2-eval-refund-3-regional-t1 | text | true | v2-eval-refund-3 | regional | 1 | 6,130,027 | 85 | refund | amount | ro-RO | null | false | 085-v2-eval-refund-3-regional-t1 | ended | false | true | false | false | 3 | 4 | true | scale | answered | false | true | 1,178 | 200 | 1 | 0 | 0 | 0 | 4,389.269 | |
text:086-v2-eval-workshop-7-explicit-t1 | text | true | v2-eval-workshop-7 | explicit | 1 | 6,142,027 | 86 | workshop | date | en-US | false | true | 086-v2-eval-workshop-7-explicit-t1 | ended | true | false | false | false | 0 | 4 | true | rank | answered | true | true | 1,025 | 206 | 1 | 0 | 0 | 0 | 4,382.8239 | |
text:087-v2-eval-workshop-7-regional-t1 | text | true | v2-eval-workshop-7 | regional | 1 | 6,142,027 | 87 | workshop | date | en-US | false | true | 087-v2-eval-workshop-7-regional-t1 | ended | true | false | false | false | 0 | 4 | true | rank | answered | true | true | 1,037 | 203 | 1 | 0 | 0 | 0 | 6,947.9271 | |
text:088-v2-eval-workshop-8-regional-t3 | text | true | v2-eval-workshop-8 | regional | 3 | 6,143,029 | 88 | workshop | date | ro-RO | false | true | 088-v2-eval-workshop-8-regional-t3 | ended | true | false | false | false | 0 | 4 | true | rank | answered | true | true | 1,037 | 274 | 1 | 0 | 0 | 0 | 5,683.3514 | |
text:089-v2-eval-workshop-8-explicit-t3 | text | true | v2-eval-workshop-8 | explicit | 3 | 6,143,029 | 89 | workshop | date | ro-RO | false | true | 089-v2-eval-workshop-8-explicit-t3 | ended | true | false | false | false | 0 | 4 | true | rank | answered | true | true | 1,025 | 243 | 1 | 0 | 0 | 0 | 5,063.3787 | |
text:090-v2-eval-purchase-5-regional-t2 | text | true | v2-eval-purchase-5 | regional | 2 | 6,124,028 | 90 | purchase | amount | en-US | null | true | 090-v2-eval-purchase-5-regional-t2 | ended | true | false | false | false | 0 | 4 | true | rank | answered | true | true | 1,124 | 250 | 1 | 0 | 0 | 0 | 5,243.9385 | |
text:091-v2-eval-purchase-5-explicit-t2 | text | true | v2-eval-purchase-5 | explicit | 2 | 6,124,028 | 91 | purchase | amount | en-US | null | true | 091-v2-eval-purchase-5-explicit-t2 | ended | true | false | false | false | 0 | 4 | true | rank | answered | true | true | 1,119 | 239 | 1 | 0 | 0 | 0 | 5,052.8382 | |
text:092-v2-eval-reschedule-1-explicit-t1 | text | true | v2-eval-reschedule-1 | explicit | 1 | 6,144,027 | 92 | reschedule | date | ro-RO | true | false | 092-v2-eval-reschedule-1-explicit-t1 | ended | true | false | false | false | 0 | 4 | true | lookup | answered | true | true | 1,094 | 194 | 1 | 0 | 0 | 0 | 4,166.8806 | |
text:093-v2-eval-reschedule-1-regional-t1 | text | true | v2-eval-reschedule-1 | regional | 1 | 6,144,027 | 93 | reschedule | date | ro-RO | true | false | 093-v2-eval-reschedule-1-regional-t1 | ended | true | false | false | false | 0 | 4 | true | lookup | answered | true | true | 1,110 | 208 | 1 | 0 | 0 | 0 | 4,411.8694 | |
text:094-v2-eval-workshop-6-regional-t1 | text | true | v2-eval-workshop-6 | regional | 1 | 6,141,027 | 94 | workshop | date | ro-RO | true | true | 094-v2-eval-workshop-6-regional-t1 | ended | true | false | false | false | 0 | 4 | true | rank | answered | true | true | 1,037 | 248 | 1 | 0 | 0 | 0 | 8,156.3069 | |
text:095-v2-eval-workshop-6-explicit-t1 | text | true | v2-eval-workshop-6 | explicit | 1 | 6,141,027 | 95 | workshop | date | ro-RO | true | true | 095-v2-eval-workshop-6-explicit-t1 | ended | true | false | false | false | 0 | 4 | true | rank | answered | true | true | 1,020 | 244 | 1 | 0 | 0 | 0 | 7,729.5359 | |
text:096-v2-eval-workshop-2-regional-t1 | text | true | v2-eval-workshop-2 | regional | 1 | 6,137,027 | 96 | workshop | date | ro-RO | true | false | 096-v2-eval-workshop-2-regional-t1 | ended | true | false | false | false | 0 | 4 | true | rank | answered | true | true | 1,037 | 254 | 1 | 0 | 0 | 0 | 8,244.4136 | |
text:097-v2-eval-workshop-2-explicit-t1 | text | true | v2-eval-workshop-2 | explicit | 1 | 6,137,027 | 97 | workshop | date | ro-RO | true | false | 097-v2-eval-workshop-2-explicit-t1 | ended | true | false | false | false | 0 | 4 | true | rank | answered | true | true | 1,020 | 209 | 1 | 0 | 0 | 0 | 6,988.5758 | |
text:098-v2-eval-reschedule-7-regional-t2 | text | true | v2-eval-reschedule-7 | regional | 2 | 6,150,028 | 98 | reschedule | date | ro-RO | false | true | 098-v2-eval-reschedule-7-regional-t2 | ended | true | false | false | false | 0 | 4 | true | lookup | answered | true | true | 1,110 | 200 | 1 | 0 | 0 | 0 | 4,203.8006 | |
text:099-v2-eval-reschedule-7-explicit-t2 | text | true | v2-eval-reschedule-7 | explicit | 2 | 6,150,028 | 99 | reschedule | date | ro-RO | false | true | 099-v2-eval-reschedule-7-explicit-t2 | ended | true | false | false | false | 0 | 4 | true | lookup | answered | true | true | 1,098 | 198 | 1 | 0 | 0 | 0 | 4,281.9496 |
Lost in Locale
Published evaluation records from an exploratory study of source formats and accepted errors in one local browser-agent policy.
Paper and full evidence archive: Zenodo, DOI 10.5281/zenodo.24058113. Author: Robert-Petre Vind-Gardoș, independent researcher. ORCID.
What is included
runs.jsonl contains 864 planned execution records, of which 862 ended.
Two user-interrupted browser runs have unobserved outcomes. Their rows remain
present with observed: false, status: incomplete and null outcome values.
They are not counted as failures or no-write outcomes.
The viewer's test split contains all three experimental conditions. It is a
publication of the held-out evaluation records, not a new random train/test
partition. Use condition, task_id, representation and trial to identify
matched cells. A task appears repeatedly across conditions, representations and
trials; rows are not 864 independent tasks.
| Condition | Planned | Ended | Unobserved |
|---|---|---|---|
text |
288 | 288 | 0 |
baseline-browser |
288 | 287 | 1 |
preview-browser |
288 | 287 | 1 |
There are 48 synthetic evaluation cases, eight in each of six workflow families: payment creation, constrained purchase, refund amendment, constrained booking, rescheduling and cancellation. Each has regional and explicit source representations and three trials. Source conventions are en-US and ro-RO; instructions and interface labels stay in English. This is not a multilingual translation benchmark.
The evaluated policy uses local qwen3:4b, Ollama 0.40.15, a model-generated
normalisation table, a deterministic calculator and accessibility-based browser
actions. The preview condition discloses the selected record's canonical source
value during review. See the paper for complete settings and development gates.
Outcomes and interpretation
For browser conditions, success means exact completion and wrong identifies
an accepted semantic error. no_write and duplicate are reported separately.
For text, these are answer-evaluation fields: text answers are not browser
writes. Diagnostic flags must not be interpreted as independent causal effects.
The paper's primary paired, equally family-weighted baseline contrast is
explicit minus regional wrong-operation rate: -7.14 percentage points, with
an exploratory 95% interval [-20.14, +2.78], family sign-flip p = 0.25.
This estimate uses 47 complete task pairs and is not the unpaired difference of
the raw arm rates. The preview supplied no observed reduction in wrong writes
on complete pairs. These results establish neither a population benefit nor
equivalence. reviewed-summary.json preserves the reviewed report verbatim.
Scope and limitations
- One qualified local policy and six related constructed workflow templates.
- Conditions were collected serially, not interleaved throughout collection.
- Two outcomes are unobserved; missingness is not assumed random.
- Text and browser conditions have different feedback/correction opportunities.
- Only the protocol 0.2 Ollama 0.40.15 cohort is included. The older pilot, failed development candidates and partial 0.40.14 cohort are not pooled.
- This repository contains derived per-run results and summaries, not the executable benchmark, original prompts, request chains or screenshots. The complete source, frozen manifests, raw logs and technical report are in the Zenodo evidence archive.
- Results on synthetic operations do not establish production safety. No independent peer-reviewed acceptance is claimed.
Reuse
from datasets import load_dataset
runs = load_dataset("robertvindgardos/lost-in-locale", split="test")
baseline = runs.filter(lambda row: row["condition"] == "baseline-browser")
observed = baseline.filter(lambda row: row["observed"])
Keep unobserved rows visible when reporting coverage. For paired estimators,
apply the complete-task rules in the paper rather than treating all rows as
independent. DATA_DICTIONARY.md describes the exported fields.
Provenance and licensing
The export checks the input CSV hashes recorded by the reviewed report, preserves every planned row, and reconciles all condition/representation outcome counts against that report. Export changes CSV strings into typed JSON values; blank inapplicable or unobserved values become null. No imputation, model run, new inferential analysis or raw-log rewrite is performed.
Collection source: 4b04c9bdea05715cc2453cc0cd138ad9da30a325.
See provenance.json and SHA256SUMS.txt for input and output hashes.
The data and documentation in this repository retain the existing MIT licence. The article copies submitted to HAL and TMLR use the separately authorised CC BY 4.0 licence; this dataset repository does not relicense those article copies. All transactions and entities are synthetic. OpenAI Codex assisted implementation, debugging, literature lookup, drafting and packaging; measured outputs came from the local model stated above.
Citation
@misc{vindgardos2026lostinlocale,
author = {Vind-Gardoș, Robert-Petre},
title = {Lost in Locale: Source Formats and Accepted Errors in a Local Browser Agent},
year = {2026},
doi = {10.5281/zenodo.24058113},
url = {https://doi.org/10.5281/zenodo.24058113},
note = {Exploratory preprint, version 1.0}
}
- Downloads last month
- 56