Add files using upload-large-folder tool

README.md

@@ -49,9 +49,9 @@ Rows with a measured per-row problem are listed in `reports/known_issues/`, one
 
 | issue | rows | train | validation | what | how it was found | file |
 |---|---:|---:|---:|---|---|---|
-| `plan_turn_rule_backwards` | 673 | 657 | 16 | The plan states the 90-degree turn backwards (e.g. on a right turn 'what was in front of me is now on my right') and still names the labelled side; the text is a rationale for the stored answer, not geometry. | thought_0 has a 'what was <front|left|right|behind> ... is/becomes/moves ... <side>' statement that contradicts the question's turn direction (right turn: front->left, right->front, behind->right, left->behind; mirrored for left); measured 2026-09-25; known_issues.py sha1 655e5307; export 4aaee1f rows 6c471ac0/6a8e1d28 | `reports/known_issues/plan_turn_rule_backwards.tsv` |
-| `contradicting_answers_same_images` | 15 | 15 | 0 | Another shipped row asks the same question (with the options in another order, or reworded) of the same input and target images and has the opposite answer; S10's conflict check compares the question with its options in listed order, so it did not see them. | rows grouped by (input-image md5, question text, option set) or by (input-image md5, question type, turn direction, last object phrase incl. 'on the left/right') whose answers differ; every member is listed; measured 2026-09-25; known_issues.py sha1 655e5307; export 4aaee1f rows 6c471ac0/6a8e1d28 | `reports/known_issues/contradicting_answers_same_images.tsv` |
-| `plan_repeats_donor_chain` | 1 | 1 | 0 | The donor wrote its chain twice, so the plan holds its Step 1-4 twice in a row, and the read-back ends with the answer twice ('My answer is [X]. The answer is X'). | thought_0 contains 'Step 1:' more than once; measured 2026-09-25; known_issues.py sha1 655e5307; export 4aaee1f rows 6c471ac0/6a8e1d28 | `reports/known_issues/plan_repeats_donor_chain.tsv` |
+| `plan_turn_rule_backwards` | 673 | 657 | 16 | The plan states the 90-degree turn backwards (e.g. on a right turn 'what was in front of me is now on my right') and still names the labelled side; the text is a rationale for the stored answer, not geometry. | thought_0 has a 'what was <front|left|right|behind> ... is/becomes/moves ... <side>' statement that contradicts the question's turn direction (right turn: front->left, right->front, behind->right, left->behind; mirrored for left); measured 2026-10-01; known_issues.py sha1 b2249331; export b3fa1b9 rows 6c471ac0/6a8e1d28 | `reports/known_issues/plan_turn_rule_backwards.tsv` |
+| `contradicting_answers_same_images` | 15 | 15 | 0 | Another shipped row asks the same question (with the options in another order, or reworded) of the same input and target images and has the opposite answer; S10's conflict check compares the question with its options in listed order, so it did not see them. | rows grouped by (input-image md5, question text, option set) or by (input-image md5, question type, turn direction, last object phrase incl. 'on the left/right') whose answers differ; every member is listed; measured 2026-10-01; known_issues.py sha1 b2249331; export b3fa1b9 rows 6c471ac0/6a8e1d28 | `reports/known_issues/contradicting_answers_same_images.tsv` |
+| `plan_repeats_donor_chain` | 1 | 1 | 0 | The donor wrote its chain twice, so the plan holds its Step 1-4 twice in a row, and the read-back ends with the answer twice ('My answer is [X]. The answer is X'). | thought_0 contains 'Step 1:' more than once; measured 2026-10-01; known_issues.py sha1 b2249331; export b3fa1b9 rows 6c471ac0/6a8e1d28 | `reports/known_issues/plan_repeats_donor_chain.tsv` |
 
 To leave the listed rows out (the snippet in the loader section downloads `reports/known_issues/` with the data):
 
@@ -251,6 +251,6 @@ Both upstream repositories were released with Bigverdi et al., *Imaginative Perc
 
 ## Provenance
 
-The release files were written by our conversion code (the code repository is not public yet), `scripts/convert/export.py` at commit `4aaee1f4e946`, from build `ipt_pet_r2`. The build was made by `scripts/convert/run_source.py` from the same repository at commit `949af62f8ab7`. S10, S12 and S13 ran before the export; `reports/export_manifest.json` pins every input the export read by SHA-1 (`build_manifest_sha1`, `s10_keep_sha1`, `s12_assignments_sha1`, `s13_balanced_keep_sha1`).
+The release files were written by our conversion code (the code repository is not public yet), `scripts/convert/export.py` at commit `b3fa1b993ca9`, from build `ipt_pet_r3`. The build was made by `scripts/convert/run_source.py` from the same repository at commit `76c78bd817c2`. S10, S12 and S13 ran before the export; `reports/export_manifest.json` pins every input the export read by SHA-1 (`build_manifest_sha1`, `s10_keep_sha1`, `s12_assignments_sha1`, `s13_balanced_keep_sha1`).
 
 Every row removed between upstream and this release has one line, with its reason, in `reports/`: `build/dropped.jsonl` (rows refused before conversion or dropped by a conversion step); `build/quarantine.jsonl` (rows set aside by S4c because an automatic check could not match the read-back's conclusion to the label); `s10_dropped.jsonl` (duplicates removed by S10); `s10_label_conflicts.jsonl` (rows S10 withheld because another row asks the identical question, options in the same order, of the same images with a different answer); `s13_dropped.jsonl` (rows removed by answer-prior balancing). `known_issues/` lists rows with a measured problem (see Known issues); `reports/` also holds the build manifest (absolute paths cut to basenames) and counters, the S14 sample list (`s14_sample.tsv`: row_uid, task, split) and `export_manifest.json`. Part of [`yrlyrl/spatial-mmcot`](https://huggingface.co/datasets/yrlyrl/spatial-mmcot).

reports/build/counters.json

@@ -10,7 +10,7 @@
    "repaired": 1,
    "quarantined": 0,
    "rejected": 0,
-   "seconds": 0.5721454620361328
+   "seconds": 0.5956151485443115
   },
   "S4c": {
    "step": "S4c",
@@ -20,7 +20,7 @@
    "repaired": 0,
    "quarantined": 0,
    "rejected": 0,
-   "seconds": 0.08992385864257812
+   "seconds": 0.09473919868469238
   },
   "S5": {
    "step": "S5",
@@ -30,7 +30,7 @@
    "repaired": 0,
    "quarantined": 0,
    "rejected": 0,
-   "seconds": 0.8670542240142822
+   "seconds": 0.8824586868286133
   },
   "S8": {
    "step": "S8",
@@ -40,7 +40,7 @@
    "repaired": 0,
    "quarantined": 0,
    "rejected": 0,
-   "seconds": 199.59299612045288
+   "seconds": 226.48321795463562
   },
   "S9": {
    "step": "S9",
@@ -50,7 +50,7 @@
    "repaired": 0,
    "quarantined": 0,
    "rejected": 0,
-   "seconds": 2.1518948078155518
+   "seconds": 2.214604139328003
   },
   "S0": {
    "step": "S0",
@@ -60,7 +60,7 @@
    "repaired": 0,
    "quarantined": 0,
    "rejected": 0,
-   "seconds": 4.887684345245361
+   "seconds": 5.036157131195068
   }
  },
  "expectations": {

reports/build/manifest.json

@@ -1,14 +1,14 @@
 {
  "slug": "ipt_pet",
  "raw": "ipt_pet",
- "out": "ipt_pet_r2",
+ "out": "ipt_pet_r3",
  "raw_rows_read": 20531,
  "raw_unusable": 0,
  "rows_out": 20531,
  "targets_out": 20531,
  "chunks": 2,
  "row_groups": 162,
- "seconds": 1787.3,
+ "seconds": 1845.1,
  "spec": {
   "slug": "ipt_pet",
   "num_inputs": 1,
@@ -26,6 +26,9 @@
   "exclude_row_uids_expected": 0,
   "exclude_image_hashes": "",
   "exclude_image_hashes_expected": 0,
+  "drop_row_uids": "",
+  "drop_row_uids_expected": 0,
+  "drop_row_uids_flag": "S5.known_bad_target",
   "crop_black_bars": false,
   "expected": {
    "S0.in_": [
@@ -36,7 +39,7 @@
  },
  "gates_failed": false,
  "limit": 0,
- "converter_git_sha": "949af62f8ab701c58da5cdeed32946242c571c83",
+ "converter_git_sha": "76c78bd817c2d8757b1929df8cf5f1a7979b143a",
  "converter_git_dirty": false,
  "paths_redacted": "absolute build-machine paths cut to basenames for publication; build_manifest_sha1 in export_manifest.json is the original's sha1"
 }

reports/export_manifest.json

@@ -1,20 +1,20 @@
 {
  "source": "ipt_pet",
- "build": "ipt_pet_r2",
- "export_git_sha": "4aaee1f4e9469da41810aac049fa411c1834b809",
+ "build": "ipt_pet_r3",
+ "export_git_sha": "b3fa1b993ca925c47fe970f2c8c727560032d7aa",
  "export_git_dirty": false,
  "limit": 0,
- "converter_git_sha": "949af62f8ab701c58da5cdeed32946242c571c83",
+ "converter_git_sha": "76c78bd817c2d8757b1929df8cf5f1a7979b143a",
  "converter_git_dirty": false,
  "build_counts": {
   "raw_rows_read": 20531,
   "raw_unusable": 0,
   "rows_out": 20531
  },
- "build_manifest_sha1": "f3af462e554778b6fe3ced7edac410575019b7c0",
- "s10_keep_sha1": "9fd27b7ff9e2be8d9e0466795f251875210de152",
- "s12_assignments_sha1": "d0195205cd5852e08769b99f9c72740f8ce9fcaa",
- "s13_balanced_keep_sha1": "af73ed4ef438de882feeee30de4a25701f2a3f6a",
+ "build_manifest_sha1": "7ad800b72ce8454e9122930601096a3e97bbd54e",
+ "s10_keep_sha1": "7a9102c5b67e340d8e47f53a99c827a78b23d530",
+ "s12_assignments_sha1": "cc2e73f7dcfb7f717a7b37bb571dec415b9f8c88",
+ "s13_balanced_keep_sha1": "87169b8ed530cb0cf4c6727b5e9b6f8a0d385f16",
  "license": "undeclared",
  "license_commercial": false,
  "upstream_repo": "weikaih/imaginative-perception-token-pet-ipt",
@@ -216,7 +216,7 @@
   "replay_unsupported": true
  },
  "s13_groups": {
-  "ipt_pet_r2/perspective_taking/train": {
+  "ipt_pet_r3/perspective_taking/train": {
    "rule": "exact_cells_pet",
    "rows_in": 19562,
    "rows_out": 8420,
@@ -247,7 +247,7 @@
    "src_split_checked": 19562,
    "src_split_disagree": 0
   },
-  "ipt_pet_r2/perspective_taking/val": {
+  "ipt_pet_r3/perspective_taking/val": {
    "rule": "exact_cells_pet",
    "rows_in": 594,
    "rows_out": 212,
@@ -280,7 +280,7 @@
   }
  },
  "blind_probe": {
-  "ipt_pet_r2/perspective_taking": {
+  "ipt_pet_r3/perspective_taking": {
    "probed": true,
    "majority": 0.2865,
    "keywords": 0.4915,

reports/known_issues/summary.json

@@ -6,7 +6,7 @@
   "validation": 0,
   "file": "known_issues/contradicting_answers_same_images.tsv",
   "detector": "rows grouped by (input-image md5, question text, option set) or by (input-image md5, question type, turn direction, last object phrase incl. 'on the left/right') whose answers differ; every member is listed",
-  "measured": "2026-09-25; known_issues.py sha1 655e5307; export 4aaee1f rows 6c471ac0/6a8e1d28"
+  "measured": "2026-10-01; known_issues.py sha1 b2249331; export b3fa1b9 rows 6c471ac0/6a8e1d28"
  },
  "plan_repeats_donor_chain": {
   "description": "The donor wrote its chain twice, so the plan holds its Step 1-4 twice in a row, and the read-back ends with the answer twice ('My answer is [X]. The answer is X').",
@@ -15,7 +15,7 @@
   "validation": 0,
   "file": "known_issues/plan_repeats_donor_chain.tsv",
   "detector": "thought_0 contains 'Step 1:' more than once",
-  "measured": "2026-09-25; known_issues.py sha1 655e5307; export 4aaee1f rows 6c471ac0/6a8e1d28"
+  "measured": "2026-10-01; known_issues.py sha1 b2249331; export b3fa1b9 rows 6c471ac0/6a8e1d28"
  },
  "plan_turn_rule_backwards": {
   "description": "The plan states the 90-degree turn backwards (e.g. on a right turn 'what was in front of me is now on my right') and still names the labelled side; the text is a rationale for the stored answer, not geometry.",
@@ -24,6 +24,6 @@
   "validation": 16,
   "file": "known_issues/plan_turn_rule_backwards.tsv",
   "detector": "thought_0 has a 'what was <front|left|right|behind> ... is/becomes/moves ... <side>' statement that contradicts the question's turn direction (right turn: front->left, right->front, behind->right, left->behind; mirrored for left)",
-  "measured": "2026-09-25; known_issues.py sha1 655e5307; export 4aaee1f rows 6c471ac0/6a8e1d28"
+  "measured": "2026-10-01; known_issues.py sha1 b2249331; export b3fa1b9 rows 6c471ac0/6a8e1d28"
  }
 }

reports/s10_label_conflicts.jsonl

@@ -1,6 +1,6 @@
-{"slug": "ipt_pet_r2", "row_uid": "63c61f0db68147dd", "step": "S10", "flag": "S10.label_conflict", "group": ["63c61f0db68147dd", "e2df7e590ce38898"], "answers_distinct": 2}
-{"slug": "ipt_pet_r2", "row_uid": "e2df7e590ce38898", "step": "S10", "flag": "S10.label_conflict", "group": ["63c61f0db68147dd", "e2df7e590ce38898"], "answers_distinct": 2}
-{"slug": "ipt_pet_r2", "row_uid": "1def04743d4e3007", "step": "S10", "flag": "S10.label_conflict", "group": ["1def04743d4e3007", "abf121767a90b5b8"], "answers_distinct": 2}
-{"slug": "ipt_pet_r2", "row_uid": "abf121767a90b5b8", "step": "S10", "flag": "S10.label_conflict", "group": ["1def04743d4e3007", "abf121767a90b5b8"], "answers_distinct": 2}
-{"slug": "ipt_pet_r2", "row_uid": "2ad72c19a6548863", "step": "S10", "flag": "S10.label_conflict", "group": ["2ad72c19a6548863", "ba44f3b042b3b483"], "answers_distinct": 2}
-{"slug": "ipt_pet_r2", "row_uid": "ba44f3b042b3b483", "step": "S10", "flag": "S10.label_conflict", "group": ["2ad72c19a6548863", "ba44f3b042b3b483"], "answers_distinct": 2}
+{"slug": "ipt_pet_r3", "row_uid": "63c61f0db68147dd", "step": "S10", "flag": "S10.label_conflict", "group": ["63c61f0db68147dd", "e2df7e590ce38898"], "answers_distinct": 2}
+{"slug": "ipt_pet_r3", "row_uid": "e2df7e590ce38898", "step": "S10", "flag": "S10.label_conflict", "group": ["63c61f0db68147dd", "e2df7e590ce38898"], "answers_distinct": 2}
+{"slug": "ipt_pet_r3", "row_uid": "1def04743d4e3007", "step": "S10", "flag": "S10.label_conflict", "group": ["1def04743d4e3007", "abf121767a90b5b8"], "answers_distinct": 2}
+{"slug": "ipt_pet_r3", "row_uid": "abf121767a90b5b8", "step": "S10", "flag": "S10.label_conflict", "group": ["1def04743d4e3007", "abf121767a90b5b8"], "answers_distinct": 2}
+{"slug": "ipt_pet_r3", "row_uid": "2ad72c19a6548863", "step": "S10", "flag": "S10.label_conflict", "group": ["2ad72c19a6548863", "ba44f3b042b3b483"], "answers_distinct": 2}
+{"slug": "ipt_pet_r3", "row_uid": "ba44f3b042b3b483", "step": "S10", "flag": "S10.label_conflict", "group": ["2ad72c19a6548863", "ba44f3b042b3b483"], "answers_distinct": 2}