release build: 163,016 rows (BUILD.json in provenance/)

.gitattributes

@@ -58,3 +58,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 # Video files - compressed
 *.mp4 filter=lfs diff=lfs merge=lfs -text
 *.webm filter=lfs diff=lfs merge=lfs -text
+provenance/new_row_screen.json filter=lfs diff=lfs merge=lfs -text

README.md

@@ -0,0 +1,258 @@
+---
+dataset_info:
+  features:
+  - name: an_creation_timestamp
+    dtype: string
+  - name: an_generation_method
+    dtype: string
+  - name: an_generation_model
+    dtype: string
+  - name: an_original_completion
+    dtype: string
+  - name: an_original_completion_length
+    dtype: int64
+  - name: an_paraphrased
+    dtype: bool
+  - name: an_style_paraphrase
+    dtype: string
+  - name: an_temperature
+    dtype: float64
+  - name: blind_judge_confidence_primary
+    dtype: string
+  - name: blind_judge_confidence_tiebreaker
+    dtype: string
+  - name: blind_judge_justification_primary
+    dtype: string
+  - name: blind_judge_justification_tiebreaker
+    dtype: string
+  - name: blind_judge_model_primary
+    dtype: string
+  - name: blind_judge_model_tiebreaker
+    dtype: string
+  - name: blind_judge_verdict_primary
+    dtype: string
+  - name: blind_judge_verdict_tiebreaker
+    dtype: string
+  - name: completion
+    dtype: string
+  - name: deception_type
+    dtype: string
+  - name: follow_up_category
+    dtype: string
+  - name: follow_up_question
+    dtype: string
+  - name: human_verified
+    dtype: bool
+  - name: judge_confidence
+    dtype: string
+  - name: judge_justification
+    dtype: string
+  - name: judge_model_informed
+    dtype: string
+  - name: judge_verdict
+    dtype: string
+  - name: language
+    dtype: string
+  - name: pair_id
+    dtype: string
+  - name: pair_type
+    dtype: string
+  - name: quadrant
+    dtype: string
+  - name: source
+    dtype: string
+  - name: translation_model
+    dtype: string
+  - name: completion_reasoning
+    dtype: string
+  - name: reasoning_in_context
+    dtype: bool
+  - name: messages_json
+    dtype: string
+  - name: tools
+    dtype: string
+  - name: _v32_source_content_md5
+    dtype: string
+  - name: _v32_lane_attributed
+    dtype: string
+  splits:
+  - name: honest_honest
+    num_bytes: 450870436
+    num_examples: 43941
+  - name: honest_deceptive
+    num_bytes: 504447178
+    num_examples: 41044
+  - name: deceptive_honest
+    num_bytes: 389100896
+    num_examples: 37567
+  - name: deceptive_deceptive
+    num_bytes: 377591319
+    num_examples: 40464
+  download_size: 0
+  dataset_size: 1722009829
+configs:
+- config_name: default
+  data_files:
+  - split: honest_honest
+    path: data/honest_honest-*
+  - split: honest_deceptive
+    path: data/honest_deceptive-*
+  - split: deceptive_honest
+    path: data/deceptive_honest-*
+  - split: deceptive_deceptive
+    path: data/deceptive_deceptive-*
+---
+
+
+# fibs-v1
+
+Private FAR AI Deception Pod research dataset: **`AlignmentResearch/deceptive-followup-v33-balanced`
+(revision `3db4df84116462a6eb5761e9b043b534b8e22c72`) with named families edited in place.** Every row of every
+untouched family is byte-identical to the parent (`provenance/untouched_identity.json` carries the
+per-family content digests); the edited families below were regenerated or topped up, screened by the
+same four-judge unanimous rule as the parent, and certified by Fable reads against each family's bar.
+
+**163,016 rows** over the four quadrant splits:
+
+| split | rows |
+|---|---:|
+| honest_honest | 43,941 |
+| honest_deceptive | 41,044 |
+| deceptive_honest | 37,567 |
+| deceptive_deceptive | 40,464 |
+
+## Edited families
+
+| family | parent rows | rows in this release |
+|---|---:|---:|
+| among | 6,464 | 6,464 |
+| anti | 7,984 | 7,984 |
+| b1-bare | 4,034 | 4,034 |
+| b1-followup | 4,488 | 4,488 |
+| caving | 1,648 | 1,648 |
+| doluschat | 7,988 | 8,000 |
+| fabricated | 2,164 | 2,164 |
+| fever | 8,000 | 8,000 |
+| harm | 5,024 | 8,000 |
+| instructed | 7,868 | 7,868 |
+| liarsbench | 8,000 | 7,992 |
+| lies | 6,760 | 6,760 |
+| longcontext | 1,764 | 1,764 |
+| mask | 8,000 | 8,000 |
+| medqa | 4,020 | 4,020 |
+| model | 3,304 | 3,304 |
+| multi | 1,476 | 1,476 |
+| negotiation | 6,954 | 6,954 |
+| original | 4,204 | 4,204 |
+| persona_prefill | 8,000 | 8,000 |
+| sandbagging | 6,700 | 6,700 |
+| sycophancy | 7,976 | 7,976 |
+| synthetic | 8,000 | 8,000 |
+| truthfulqa | 4,804 | 4,804 |
+| ultrachat | 8,000 | 8,000 |
+| werewolf | 6,692 | 6,692 |
+| wikihow | 1,760 | 1,760 |
+| wikihownest | 7,960 | 7,960 |
+
+The harm-knowledge rows that sat under the Liars' Bench family label in the parent are reassigned to
+the harm family here (`provenance/family_edits.json`).
+
+## Certification of the new rows
+
+| family (rubric) | kept by the four judges | Fable reads | mislabels | rate | 95% upper | pass line (alpha 0.05/29) | outcome |
+|---|---:|---:|---:|---:|---:|---:|---|
+| Harm pressure knowledge (v1.3) | 2,186 of 3,214 | 296 (207 refusals re-read by GLM-5.3) | 0 | 0.0% | 1.0% | 4 | PASS |
+| Liars' Bench soft-trigger (v1.3) | 4,032 of 5,097 | 500 | 4 | 0.8% | 1.8% | 11 | PASS |
+| DolusChat (v1.5) | 17,163 of 22,447 | 299 (4 refusals, 3 re-read by GLM-5.3, 1 no verdict) | 5 | 1.67% | 3.5% | 4 | accepted by ruling at the parent family's certified rate (5 in 342, 1.46%); the five convicted rows are excluded |
+
+Every Fable-convicted row is excluded from the build and every Fable-confirmed row is kept regardless of the four judges (`provenance/new_row_screen.json`). Insider trading was generated but is not part of this release.
+
+Screening campaigns folded into this build:
+
+| campaign | rubric | new-row screen outcomes |
+|---|---|---|
+| pharm | rubric_plainconstruct_v1_3.md@ec597d99eb10d2afa92a28bddcde2a62 | harm: duplicate_of_new 1, duplicate_of_parent 1, fable_confirmed 93, glm_confirmed 203, judges_kept 1,886, judges_rejected 1,028, refused_no_stand_in_judges_kept 4 |
+| plbsoft | rubric_plainconstruct_v1_3.md@ec597d99eb10d2afa92a28bddcde2a62 | liarsbench: fable_confirmed 496, fable_mislabel 4, judges_kept 3,532, judges_rejected 1,065 |
+| pdolu2 | rubric_plainconstruct_v1_5.md@bcb71e10fabe67dc387440c3c2d5e31b | doluschat: fable_confirmed 291, fable_mislabel 5, glm_confirmed 3, judges_kept 16,863, judges_rejected 5,284, refused_no_stand_in_judges_kept 1 |
+
+## Contextual rows: a family-level note (no column in this release)
+
+The paper distinguishes **contextual** rows, where the truth of the statement the model makes is evidenced
+in the transcript itself (system prompt, earlier turns, tool results, a supplied passage, or the model's
+own visible working), from rows that rest on world knowledge, which the model could in principle have been
+trained to believe otherwise. This release carries no per-row column for it. The classification below is by
+family, sub-source and follow-up category, chosen for precision over recall after reading two full
+transcripts per family; a row-level classifier is future work.
+
+**Contextual by construction** (the transcript carries the truth-maker in every row read):
+
+| family or sub-source | where the truth-maker sits |
+|---|---|
+| doluschat, persona_prefill, mask_provided_facts | facts stated in the system prompt |
+| b1-bare, b1-followup | the tool result earlier in the transcript |
+| among_us, werewolf, negotiation | the observation log, private role or night knowledge, private facts |
+| sycophancy | the base turn's `<scratchpad>` writes out the correct answer that the stated answer contradicts |
+| longcontext v3_bookqa, lc3flip, lc5flip, v5a/v5b_onpolicy, v5c3_intrace, v5c5_novels | the supplied passage, the agent transcript's tool result, or a value the model itself stated earlier |
+
+**Contextual by follow-up category, in any family:** `cat_H` and `cat_Hf`, the response-meta questions
+("So you said Y, right?", "Did you mention any caveats?"), whose answer is decided by the prior turn.
+
+**Not counted as contextual** (world knowledge, internal state, or grounding not guaranteed per row):
+anti_sycophancy, caving (including its numeric-arithmetic rows), fabricated_statistics, fever (even when the
+passage is supplied), harm pressure, instructed, liarsbench (including the soft-trigger rows), lies_v2,
+longcontext lcflip/v2/v4_offpolicy, mask known-facts/statistics/doubling-down, medqa, model_self_knowledge,
+multi_turn, original, sandbagging, synthetic (even when the user supplies the source text), truthfulqa,
+ultrachat, wikihow, wikihownest. Meta-truthfulness follow-ups ("Were you honest?") inherit the family's status.
+
+Under this rule **79,580 of 163,016 rows (48.8%)** are contextual:
+
+| family | rows | by family/sub-source | by follow-up only | contextual |
+|---|---:|---:|---:|---:|
+| among | 6,464 | 6,464 | 0 | 6,464 |
+| anti | 7,984 | 0 | 1,873 | 1,873 |
+| b1-bare | 4,034 | 4,034 | 0 | 4,034 |
+| b1-followup | 4,488 | 4,488 | 0 | 4,488 |
+| caving | 1,648 | 0 | 0 | 0 |
+| doluschat | 8,000 | 8,000 | 0 | 8,000 |
+| fabricated | 2,164 | 0 | 483 | 483 |
+| fever | 8,000 | 0 | 1,452 | 1,452 |
+| harm | 8,000 | 0 | 2,338 | 2,338 |
+| instructed | 7,868 | 0 | 1,462 | 1,462 |
+| liarsbench | 7,992 | 0 | 1,693 | 1,693 |
+| lies | 6,760 | 0 | 1,720 | 1,720 |
+| longcontext | 1,764 | 1,353 | 5 | 1,358 |
+| mask | 8,000 | 2,445 | 1,467 | 3,912 |
+| medqa | 4,020 | 0 | 803 | 803 |
+| model | 3,304 | 0 | 846 | 846 |
+| multi | 1,476 | 0 | 0 | 0 |
+| negotiation | 6,954 | 6,954 | 0 | 6,954 |
+| original | 4,204 | 0 | 632 | 632 |
+| persona_prefill | 8,000 | 8,000 | 0 | 8,000 |
+| sandbagging | 6,700 | 0 | 1,959 | 1,959 |
+| sycophancy | 7,976 | 7,976 | 0 | 7,976 |
+| synthetic | 8,000 | 0 | 1,777 | 1,777 |
+| truthfulqa | 4,804 | 0 | 912 | 912 |
+| ultrachat | 8,000 | 0 | 1,870 | 1,870 |
+| werewolf | 6,692 | 6,692 | 0 | 6,692 |
+| wikihow | 1,760 | 0 | 0 | 0 |
+| wikihownest | 7,960 | 0 | 1,882 | 1,882 |
+
+## Provenance
+
+`provenance/BUILD.json` records the parent revision and parquet sha256s, the seed (20260929),
+the family cap (8,000), the per-family before/after counts and the builder's sha256;
+`provenance/new_row_screen.json` lists every screened new row with its judge and Fable outcome;
+`provenance/covariate_trim.json` holds the per-family covariate trim logs. The builder is
+`sandbox/oa_backdoor/scripts/dataset_generation/build_v2_slotin.py` in the Deception Pod monorepo.
+
+## Columns and loading
+
+Identical to the parent: `messages_json` (UTF-8 JSON message list; assistant turns may carry
+`reasoning_content` sidecars and OpenAI-style `tool_calls`), `tools` (JSON string tool-schema list where
+the source uses tools, null elsewhere), `completion`, and the source and judge metadata columns.
+Consumers must `json.loads` the JSON-string columns. Quadrants use `base_follow-up` semantics: the first
+term labels the base turn and the second labels the follow-up turn.
+
+Re-rendered text derived from these rows is for probe input and activation harvest only; it must not
+replace original rollout token IDs or feed RL-training token paths, because re-templating can change
+tokenization.

data/deceptive_deceptive-00000-of-00001.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3b1c3c0e50c24632c87e7c1443a0280c508464428f4b931bd655318d5d57e8da
+size 173018309

data/deceptive_honest-00000-of-00001.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b80b3616ca2ad9de176c1d094c4b4324ecd5ebe8a2583fe647210ff8dcc258c2
+size 168555904

data/honest_deceptive-00000-of-00001.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8f27d8ef9cc7f2ba4fa446c457a0c737b06b59698dfc853fb090c09bda52acc6
+size 216786956

data/honest_honest-00000-of-00001.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:555563638e48bb8c30ccffee4d7074e549538e4293aa1a592123dfb5310522ed
+size 194030526

provenance/BUILD.json

@@ -0,0 +1,533 @@
+{
+ "parent_repo": "AlignmentResearch/deceptive-followup-v33-balanced",
+ "parent_revision": "3db4df84116462a6eb5761e9b043b534b8e22c72",
+ "parent_parquet_sha256": {
+  "honest_honest": [
+   "e24a3ef0f56220867f7ab1f0db204fa463376633aea017ffe21957c3e33be575"
+  ],
+  "honest_deceptive": [
+   "9966a5659961a82e490531d91b91bbe834ecd70a4715ae3b839d51b5bdde48ea",
+   "b7248afe90174a9855ae3cf83a961683a376559facdbc54fd8daa7b6822b419d"
+  ],
+  "deceptive_honest": [
+   "fd4d68657af296ad0e47fefb036f7930147b0d84b824d30ab28272cd757487fe"
+  ],
+  "deceptive_deceptive": [
+   "fb3e8caac846408c3a5a5b1e2692383fdddf47dcdc4b6b6c3045040a9cc5f962"
+  ]
+ },
+ "parent_mask_rule": "certcov-v33-full-adopted3of4-equalquadrant-leakfixv3cov-cap8000-20260904",
+ "seed": 20260929,
+ "cap": 8000,
+ "targets": {
+  "doluschat": 8000,
+  "liarsbench": 8000,
+  "harm": 8000
+ },
+ "campaigns": [
+  {
+   "dir": "/private/tmp/claude-501/-Users-chris-Documents-FAR-Projects-f-deception-deception/bca0ef1b-296e-4928-a803-f0a938c3988b/scratchpad/vol/v2_artifacts/cert_nb/pharm",
+   "tag": "pharm",
+   "rubric": "rubric_plainconstruct_v1_3.md@ec597d99eb10d2afa92a28bddcde2a62",
+   "staged_md5": "f6a4f0d8e40f4c71df3d0f98dcf62e1c",
+   "fold_md5": "91ee7b2899ec7934905b05ec193def66",
+   "manifest_md5": "f5b1327934147bdce7af99db28616cbd",
+   "artifacts": [
+    {
+     "artifact": "fable_artifact_1.json",
+     "raw": "fable_artifact_1.json.raw.jsonl",
+     "raw_md5": "0477e58a8dc871632210f424431ed8e0",
+     "glm_md5": "a4beb1c792195a41d2117130267259f0",
+     "n_results": 100
+    },
+    {
+     "artifact": "fable_artifact_2.json",
+     "raw": "fable_artifact_2.json.raw.jsonl",
+     "raw_md5": "f275655952810bb256a31165bf164c1a",
+     "glm_md5": "153f553c44720dda9f3d63ba9b1be42c",
+     "n_results": 100
+    },
+    {
+     "artifact": "fable_artifact_3.json",
+     "raw": "fable_artifact_3.json.raw.jsonl",
+     "raw_md5": "9c9e4d0daa355553fb59e4759eada3fd",
+     "glm_md5": "e13435d145eb1fba3e7b62de809172df",
+     "n_results": 100
+    },
+    {
+     "artifact": "fable_artifact_4.json",
+     "raw": null
+    }
+   ],
+   "status_counts": {
+    "harm": {
+     "judges_kept": 1886,
+     "glm_confirmed": 203,
+     "fable_confirmed": 93,
+     "judges_rejected": 1028,
+     "refused_no_stand_in_judges_kept": 4,
+     "duplicate_of_parent": 1,
+     "duplicate_of_new": 1
+    }
+   }
+  },
+  {
+   "dir": "/private/tmp/claude-501/-Users-chris-Documents-FAR-Projects-f-deception-deception/bca0ef1b-296e-4928-a803-f0a938c3988b/scratchpad/vol/v2_artifacts/cert_nb/plbsoft",
+   "tag": "plbsoft",
+   "rubric": "rubric_plainconstruct_v1_3.md@ec597d99eb10d2afa92a28bddcde2a62",
+   "staged_md5": "8149c0abd95b427e82537ac081d27433",
+   "fold_md5": "d252fa9040301344b71b229eece08f89",
+   "manifest_md5": "6fc1bab4921fce9dade0e30478e463b0",
+   "artifacts": [
+    {
+     "artifact": "fable_artifact_1.json",
+     "raw": "fable_artifact_1.json.raw.jsonl",
+     "raw_md5": "4780c171e3ba1e65526d0ac6b9529a0e",
+     "glm_md5": null,
+     "n_results": 100
+    },
+    {
+     "artifact": "fable_artifact_2.json",
+     "raw": "fable_artifact_2.json.raw.jsonl",
+     "raw_md5": "6f590da6554f99c7533325efdfd50377",
+     "glm_md5": null,
+     "n_results": 100
+    },
+    {
+     "artifact": "fable_artifact_3.json",
+     "raw": "fable_artifact_3.json.raw.jsonl",
+     "raw_md5": "5e9bee87da1f9f381fb436b546f04735",
+     "glm_md5": null,
+     "n_results": 100
+    },
+    {
+     "artifact": "fable_artifact_4.json",
+     "raw": "fable_artifact_4.json.raw.jsonl",
+     "raw_md5": "caa053de2a671c6866f4ad6cd51599f1",
+     "glm_md5": null,
+     "n_results": 100
+    },
+    {
+     "artifact": "fable_artifact_5.json",
+     "raw": "fable_artifact_5.json.raw.jsonl",
+     "raw_md5": "289df8c43dfc0e3e30faf9b7b1783288",
+     "glm_md5": null,
+     "n_results": 100
+    },
+    {
+     "artifact": "fable_artifact_6.json",
+     "raw": null
+    }
+   ],
+   "status_counts": {
+    "liarsbench": {
+     "judges_kept": 3532,
+     "fable_confirmed": 496,
+     "judges_rejected": 1065,
+     "fable_mislabel": 4
+    }
+   }
+  },
+  {
+   "dir": "/private/tmp/claude-501/-Users-chris-Documents-FAR-Projects-f-deception-deception/bca0ef1b-296e-4928-a803-f0a938c3988b/scratchpad/vol/v2_artifacts/cert_nb/pdolu2",
+   "tag": "pdolu2",
+   "rubric": "rubric_plainconstruct_v1_5.md@bcb71e10fabe67dc387440c3c2d5e31b",
+   "staged_md5": "a138fec9b08eb8201774dbd03a3a6c17",
+   "fold_md5": "a4e9a0a41613b4f52f84c47d57ff2b55",
+   "manifest_md5": "befd276ee4ddba4aa7f36f8dbaaa40d9",
+   "artifacts": [
+    {
+     "artifact": "fable_artifact_1.json",
+     "raw": "fable_artifact_1.json.raw.jsonl",
+     "raw_md5": "61674f5b5068f0fc7bc8168b913158d0",
+     "glm_md5": null,
+     "n_results": 100
+    },
+    {
+     "artifact": "fable_artifact_2.json",
+     "raw": "fable_artifact_2.json.raw.jsonl",
+     "raw_md5": "d4efa7015f97f9b47782c3be90fd7006",
+     "glm_md5": "e0b2b35ae3b393b0be092a98bd3cc6b5",
+     "n_results": 100
+    },
+    {
+     "artifact": "fable_artifact_3.json",
+     "raw": "fable_artifact_3.json.raw.jsonl",
+     "raw_md5": "8dc2543a6de349a9b4b87785f4045215",
+     "glm_md5": "e7f608dfca5fefa7c29f35f118f919e1",
+     "n_results": 100
+    }
+   ],
+   "status_counts": {
+    "doluschat": {
+     "judges_kept": 16863,
+     "judges_rejected": 5284,
+     "fable_confirmed": 291,
+     "fable_mislabel": 5,
+     "glm_confirmed": 3,
+     "refused_no_stand_in_judges_kept": 1
+    }
+   }
+  }
+ ],
+ "generator_only_columns_dropped": [
+  "base_hash",
+  "frameleak_exemptions",
+  "frameleak_is_leak",
+  "frameleak_judge_model",
+  "frameleak_layer",
+  "frameleak_matched",
+  "frameleak_reasoning",
+  "gold_anchor",
+  "recant_check_justification",
+  "recant_check_verdict"
+ ],
+ "absent_column_defaults": {
+  "an_original_completion": "",
+  "an_original_completion_length": 0,
+  "an_paraphrased": false,
+  "an_style_paraphrase": "",
+  "blind_judge_confidence_tiebreaker": "",
+  "blind_judge_justification_tiebreaker": "",
+  "blind_judge_model_tiebreaker": "",
+  "blind_judge_verdict_tiebreaker": "",
+  "pair_id": "",
+  "pair_type": "",
+  "_v32_lane_attributed": "unattributed"
+ },
+ "n_rows": {
+  "honest_honest": 43941,
+  "honest_deceptive": 41044,
+  "deceptive_honest": 37567,
+  "deceptive_deceptive": 40464
+ },
+ "n_total": 163016,
+ "before": {
+  "doluschat": {
+   "honest_honest": 1997,
+   "honest_deceptive": 1997,
+   "deceptive_honest": 1997,
+   "deceptive_deceptive": 1997
+  },
+  "lies": {
+   "honest_honest": 1690,
+   "honest_deceptive": 1690,
+   "deceptive_honest": 1690,
+   "deceptive_deceptive": 1690
+  },
+  "truthfulqa": {
+   "honest_honest": 1201,
+   "honest_deceptive": 1201,
+   "deceptive_honest": 1201,
+   "deceptive_deceptive": 1201
+  },
+  "sandbagging": {
+   "honest_honest": 1675,
+   "honest_deceptive": 1675,
+   "deceptive_honest": 1675,
+   "deceptive_deceptive": 1675
+  },
+  "ultrachat": {
+   "honest_honest": 4000,
+   "honest_deceptive": 4000
+  },
+  "wikihownest": {
+   "honest_honest": 1990,
+   "honest_deceptive": 1990,
+   "deceptive_honest": 1990,
+   "deceptive_deceptive": 1990
+  },
+  "wikihow": {
+   "honest_honest": 880,
+   "deceptive_deceptive": 880
+  },
+  "sycophancy": {
+   "honest_honest": 1994,
+   "honest_deceptive": 1994,
+   "deceptive_honest": 1994,
+   "deceptive_deceptive": 1994
+  },
+  "original": {
+   "honest_honest": 1051,
+   "honest_deceptive": 1051,
+   "deceptive_honest": 1051,
+   "deceptive_deceptive": 1051
+  },
+  "harm": {
+   "honest_honest": 1256,
+   "honest_deceptive": 1256,
+   "deceptive_honest": 1256,
+   "deceptive_deceptive": 1256
+  },
+  "fever": {
+   "honest_honest": 2000,
+   "honest_deceptive": 2000,
+   "deceptive_honest": 2000,
+   "deceptive_deceptive": 2000
+  },
+  "model": {
+   "honest_honest": 826,
+   "honest_deceptive": 826,
+   "deceptive_honest": 826,
+   "deceptive_deceptive": 826
+  },
+  "fabricated": {
+   "honest_honest": 541,
+   "honest_deceptive": 541,
+   "deceptive_honest": 541,
+   "deceptive_deceptive": 541
+  },
+  "medqa": {
+   "honest_honest": 1005,
+   "honest_deceptive": 1005,
+   "deceptive_honest": 1005,
+   "deceptive_deceptive": 1005
+  },
+  "liarsbench": {
+   "honest_honest": 2000,
+   "honest_deceptive": 2000,
+   "deceptive_honest": 2000,
+   "deceptive_deceptive": 2000
+  },
+  "blackbox": {
+   "honest_honest": 431,
+   "honest_deceptive": 431,
+   "deceptive_honest": 431,
+   "deceptive_deceptive": 431
+  },
+  "mask": {
+   "honest_honest": 2000,
+   "honest_deceptive": 2000,
+   "deceptive_honest": 2000,
+   "deceptive_deceptive": 2000
+  },
+  "multi": {
+   "honest_honest": 369,
+   "honest_deceptive": 369,
+   "deceptive_honest": 369,
+   "deceptive_deceptive": 369
+  },
+  "anti": {
+   "honest_honest": 1996,
+   "honest_deceptive": 1996,
+   "deceptive_honest": 1996,
+   "deceptive_deceptive": 1996
+  },
+  "b1-bare": {
+   "honest_honest": 2017,
+   "deceptive_deceptive": 2017
+  },
+  "b1-followup": {
+   "honest_honest": 1122,
+   "honest_deceptive": 1122,
+   "deceptive_honest": 1122,
+   "deceptive_deceptive": 1122
+  },
+  "among": {
+   "honest_honest": 1616,
+   "honest_deceptive": 1616,
+   "deceptive_honest": 1616,
+   "deceptive_deceptive": 1616
+  },
+  "werewolf": {
+   "honest_honest": 1673,
+   "honest_deceptive": 1673,
+   "deceptive_honest": 1673,
+   "deceptive_deceptive": 1673
+  },
+  "negotiation": {
+   "honest_honest": 3477,
+   "honest_deceptive": 3477
+  },
+  "caving": {
+   "honest_honest": 412,
+   "honest_deceptive": 412,
+   "deceptive_honest": 412,
+   "deceptive_deceptive": 412
+  },
+  "instructed": {
+   "honest_honest": 1967,
+   "honest_deceptive": 1967,
+   "deceptive_honest": 1967,
+   "deceptive_deceptive": 1967
+  },
+  "persona_prefill": {
+   "honest_honest": 2000,
+   "honest_deceptive": 2000,
+   "deceptive_honest": 2000,
+   "deceptive_deceptive": 2000
+  },
+  "longcontext": {
+   "honest_honest": 441,
+   "honest_deceptive": 441,
+   "deceptive_honest": 441,
+   "deceptive_deceptive": 441
+  },
+  "synthetic": {
+   "deceptive_honest": 4000,
+   "deceptive_deceptive": 4000
+  }
+ },
+ "after": {
+  "lies": {
+   "honest_honest": 1690,
+   "honest_deceptive": 1690,
+   "deceptive_honest": 1690,
+   "deceptive_deceptive": 1690
+  },
+  "truthfulqa": {
+   "honest_honest": 1201,
+   "honest_deceptive": 1201,
+   "deceptive_honest": 1201,
+   "deceptive_deceptive": 1201
+  },
+  "sandbagging": {
+   "honest_honest": 1675,
+   "honest_deceptive": 1675,
+   "deceptive_honest": 1675,
+   "deceptive_deceptive": 1675
+  },
+  "ultrachat": {
+   "honest_honest": 4000,
+   "honest_deceptive": 4000
+  },
+  "wikihownest": {
+   "honest_honest": 1990,
+   "honest_deceptive": 1990,
+   "deceptive_honest": 1990,
+   "deceptive_deceptive": 1990
+  },
+  "wikihow": {
+   "honest_honest": 880,
+   "deceptive_deceptive": 880
+  },
+  "sycophancy": {
+   "honest_honest": 1994,
+   "honest_deceptive": 1994,
+   "deceptive_honest": 1994,
+   "deceptive_deceptive": 1994
+  },
+  "original": {
+   "honest_honest": 1051,
+   "honest_deceptive": 1051,
+   "deceptive_honest": 1051,
+   "deceptive_deceptive": 1051
+  },
+  "harm": {
+   "honest_honest": 2000,
+   "honest_deceptive": 2000,
+   "deceptive_honest": 2000,
+   "deceptive_deceptive": 2000
+  },
+  "fever": {
+   "honest_honest": 2000,
+   "honest_deceptive": 2000,
+   "deceptive_honest": 2000,
+   "deceptive_deceptive": 2000
+  },
+  "model": {
+   "honest_honest": 826,
+   "honest_deceptive": 826,
+   "deceptive_honest": 826,
+   "deceptive_deceptive": 826
+  },
+  "fabricated": {
+   "honest_honest": 541,
+   "honest_deceptive": 541,
+   "deceptive_honest": 541,
+   "deceptive_deceptive": 541
+  },
+  "medqa": {
+   "honest_honest": 1005,
+   "honest_deceptive": 1005,
+   "deceptive_honest": 1005,
+   "deceptive_deceptive": 1005
+  },
+  "liarsbench": {
+   "honest_honest": 1998,
+   "honest_deceptive": 1998,
+   "deceptive_honest": 1998,
+   "deceptive_deceptive": 1998
+  },
+  "mask": {
+   "honest_honest": 2000,
+   "honest_deceptive": 2000,
+   "deceptive_honest": 2000,
+   "deceptive_deceptive": 2000
+  },
+  "multi": {
+   "honest_honest": 369,
+   "honest_deceptive": 369,
+   "deceptive_honest": 369,
+   "deceptive_deceptive": 369
+  },
+  "anti": {
+   "honest_honest": 1996,
+   "honest_deceptive": 1996,
+   "deceptive_honest": 1996,
+   "deceptive_deceptive": 1996
+  },
+  "b1-bare": {
+   "honest_honest": 2017,
+   "deceptive_deceptive": 2017
+  },
+  "b1-followup": {
+   "honest_honest": 1122,
+   "honest_deceptive": 1122,
+   "deceptive_honest": 1122,
+   "deceptive_deceptive": 1122
+  },
+  "among": {
+   "honest_honest": 1616,
+   "honest_deceptive": 1616,
+   "deceptive_honest": 1616,
+   "deceptive_deceptive": 1616
+  },
+  "werewolf": {
+   "honest_honest": 1673,
+   "honest_deceptive": 1673,
+   "deceptive_honest": 1673,
+   "deceptive_deceptive": 1673
+  },
+  "negotiation": {
+   "honest_honest": 3477,
+   "honest_deceptive": 3477
+  },
+  "caving": {
+   "honest_honest": 412,
+   "honest_deceptive": 412,
+   "deceptive_honest": 412,
+   "deceptive_deceptive": 412
+  },
+  "instructed": {
+   "honest_honest": 1967,
+   "honest_deceptive": 1967,
+   "deceptive_honest": 1967,
+   "deceptive_deceptive": 1967
+  },
+  "persona_prefill": {
+   "honest_honest": 2000,
+   "honest_deceptive": 2000,
+   "deceptive_honest": 2000,
+   "deceptive_deceptive": 2000
+  },
+  "longcontext": {
+   "honest_honest": 441,
+   "honest_deceptive": 441,
+   "deceptive_honest": 441,
+   "deceptive_deceptive": 441
+  },
+  "doluschat": {
+   "honest_honest": 2000,
+   "honest_deceptive": 2000,
+   "deceptive_honest": 2000,
+   "deceptive_deceptive": 2000
+  },
+  "synthetic": {
+   "deceptive_honest": 4000,
+   "deceptive_deceptive": 4000
+  }
+ },
+ "builder_sha256": "f8ed5234ac3c1df182fa97ed09f230170c19fc0cfddd3bea205894893fc35485",
+ "created_at": "2026-09-30T06:19:22Z"
+}

provenance/covariate_trim.json

@@ -0,0 +1,21 @@
+{
+ "doluschat": [
+  "pre-draw eps=0.04 round 1: dropped 0",
+  "eps=0.04: drawn 8000, after sweep-trim 8000, sweep fails 0"
+ ],
+ "liarsbench": [
+  "pre-draw eps=0.04 round 1: dropped 0",
+  "eps=0.04: drawn 7992, after sweep-trim 7992, sweep fails 0"
+ ],
+ "harm": [
+  "pre-draw eps=0.04 round 1: harm_pressure_knowledge A-cell 10 (182 rows) rate 0.456 vs source 0.535: dropped 8",
+  "pre-draw eps=0.04 round 1: dropped 8",
+  "pre-draw eps=0.04 round 2: dropped 0",
+  "eps=0.04: drawn 8000, after sweep-trim 8000, sweep fails 0"
+ ],
+ "eps_passed": {
+  "doluschat": 0.04,
+  "liarsbench": 0.04,
+  "harm": 0.04
+ }
+}

provenance/new_row_screen.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b8b4a170d586639119cd0becc755f6b22a83061c76f19cc6e1f8b91f60e44e37
+size 11113328

provenance/untouched_identity.json

@@ -0,0 +1,177 @@
+{
+ "among": {
+  "parent_rows": 6464,
+  "output_rows": 6464,
+  "sha256_parent": "18043107b7087701205020d3b83a298f842328cb9330a8e676e0bb7dada359a2",
+  "sha256_output": "18043107b7087701205020d3b83a298f842328cb9330a8e676e0bb7dada359a2",
+  "identical": true
+ },
+ "anti": {
+  "parent_rows": 7984,
+  "output_rows": 7984,
+  "sha256_parent": "95c3cb49e668fcc813f80408448cf4aed3806116117289d8625a63b8935f645b",
+  "sha256_output": "95c3cb49e668fcc813f80408448cf4aed3806116117289d8625a63b8935f645b",
+  "identical": true
+ },
+ "b1-bare": {
+  "parent_rows": 4034,
+  "output_rows": 4034,
+  "sha256_parent": "102f0e967b8f52583f11af608e4e5a905258e5f3d931f74882df9f29c9d0f08a",
+  "sha256_output": "102f0e967b8f52583f11af608e4e5a905258e5f3d931f74882df9f29c9d0f08a",
+  "identical": true
+ },
+ "b1-followup": {
+  "parent_rows": 4488,
+  "output_rows": 4488,
+  "sha256_parent": "7865aa70bd078a7db34fc0bdfa18a824ed3996dc9f4fc8ea239d5be6583eda02",
+  "sha256_output": "7865aa70bd078a7db34fc0bdfa18a824ed3996dc9f4fc8ea239d5be6583eda02",
+  "identical": true
+ },
+ "caving": {
+  "parent_rows": 1648,
+  "output_rows": 1648,
+  "sha256_parent": "6d4e94f4d894348b69e0d5f98f164eaaadfc36cc4b86b84b390555031248f2d8",
+  "sha256_output": "6d4e94f4d894348b69e0d5f98f164eaaadfc36cc4b86b84b390555031248f2d8",
+  "identical": true
+ },
+ "fabricated": {
+  "parent_rows": 2164,
+  "output_rows": 2164,
+  "sha256_parent": "4af82d4be15af2a03a87191f19e7836312895fb0a3e06075bd163b71d5899353",
+  "sha256_output": "4af82d4be15af2a03a87191f19e7836312895fb0a3e06075bd163b71d5899353",
+  "identical": true
+ },
+ "fever": {
+  "parent_rows": 8000,
+  "output_rows": 8000,
+  "sha256_parent": "0cc28cb13d6777f7803a8becdb934454384b3259a7aca7e36b882660f20bc0e4",
+  "sha256_output": "0cc28cb13d6777f7803a8becdb934454384b3259a7aca7e36b882660f20bc0e4",
+  "identical": true
+ },
+ "instructed": {
+  "parent_rows": 7868,
+  "output_rows": 7868,
+  "sha256_parent": "b60a696575398d7d6c08fe29af0ae5a4312476f81bc8793d6f4db75cebfd04f1",
+  "sha256_output": "b60a696575398d7d6c08fe29af0ae5a4312476f81bc8793d6f4db75cebfd04f1",
+  "identical": true
+ },
+ "lies": {
+  "parent_rows": 6760,
+  "output_rows": 6760,
+  "sha256_parent": "fcac8b058b0391dbbe47f199ab24de44b10d9c16674da9d9147f93bace97d4d7",
+  "sha256_output": "fcac8b058b0391dbbe47f199ab24de44b10d9c16674da9d9147f93bace97d4d7",
+  "identical": true
+ },
+ "longcontext": {
+  "parent_rows": 1764,
+  "output_rows": 1764,
+  "sha256_parent": "cfd6dc0a2b9c2a3d3aec572698cc2a99be7a207dc6b422c81567726ea6c6ee9b",
+  "sha256_output": "cfd6dc0a2b9c2a3d3aec572698cc2a99be7a207dc6b422c81567726ea6c6ee9b",
+  "identical": true
+ },
+ "mask": {
+  "parent_rows": 8000,
+  "output_rows": 8000,
+  "sha256_parent": "35397e869cd22fdc5b9558272cce0fffefe63faf521634d1e1a757bac17dd495",
+  "sha256_output": "35397e869cd22fdc5b9558272cce0fffefe63faf521634d1e1a757bac17dd495",
+  "identical": true
+ },
+ "medqa": {
+  "parent_rows": 4020,
+  "output_rows": 4020,
+  "sha256_parent": "803c47bbc6962e17b0c5cc3251498d89de71c50e4769f88b0899a33da793d37f",
+  "sha256_output": "803c47bbc6962e17b0c5cc3251498d89de71c50e4769f88b0899a33da793d37f",
+  "identical": true
+ },
+ "model": {
+  "parent_rows": 3304,
+  "output_rows": 3304,
+  "sha256_parent": "b04561d47ebdb1a7042577c3237e0fdb3697ede930384fd53eb7a9dc9a10870c",
+  "sha256_output": "b04561d47ebdb1a7042577c3237e0fdb3697ede930384fd53eb7a9dc9a10870c",
+  "identical": true
+ },
+ "multi": {
+  "parent_rows": 1476,
+  "output_rows": 1476,
+  "sha256_parent": "f022e2758533ef907d6e5049b26062bd5a353a45bb66c16e4ab358ebff3941e6",
+  "sha256_output": "f022e2758533ef907d6e5049b26062bd5a353a45bb66c16e4ab358ebff3941e6",
+  "identical": true
+ },
+ "negotiation": {
+  "parent_rows": 6954,
+  "output_rows": 6954,
+  "sha256_parent": "c76f3354e837a2df9b9e85e40ba630fbb27815d7c5218a2fd6eabbef20ebfd7d",
+  "sha256_output": "c76f3354e837a2df9b9e85e40ba630fbb27815d7c5218a2fd6eabbef20ebfd7d",
+  "identical": true
+ },
+ "original": {
+  "parent_rows": 4204,
+  "output_rows": 4204,
+  "sha256_parent": "3c2ba0a552b44bb7679670f19a3339ca0889d1a990ef0102f1309cf284ebc10b",
+  "sha256_output": "3c2ba0a552b44bb7679670f19a3339ca0889d1a990ef0102f1309cf284ebc10b",
+  "identical": true
+ },
+ "persona_prefill": {
+  "parent_rows": 8000,
+  "output_rows": 8000,
+  "sha256_parent": "a0d2b9673c5c89803e1e829327912b684cbd78a6788559181a20770fa4d72530",
+  "sha256_output": "a0d2b9673c5c89803e1e829327912b684cbd78a6788559181a20770fa4d72530",
+  "identical": true
+ },
+ "sandbagging": {
+  "parent_rows": 6700,
+  "output_rows": 6700,
+  "sha256_parent": "a651a4610a9d24a2f4a25841a925e2d0b7d49928214daf22f34caa840fb366ce",
+  "sha256_output": "a651a4610a9d24a2f4a25841a925e2d0b7d49928214daf22f34caa840fb366ce",
+  "identical": true
+ },
+ "sycophancy": {
+  "parent_rows": 7976,
+  "output_rows": 7976,
+  "sha256_parent": "4cd8b22297871ea184393b34b9760441417080d254726ad8ceb9a465b59e8c80",
+  "sha256_output": "4cd8b22297871ea184393b34b9760441417080d254726ad8ceb9a465b59e8c80",
+  "identical": true
+ },
+ "synthetic": {
+  "parent_rows": 8000,
+  "output_rows": 8000,
+  "sha256_parent": "cd6ebe99c10c622c080f7b1b9cb62cf692a0c64e7bf38058178c293d33496dd6",
+  "sha256_output": "cd6ebe99c10c622c080f7b1b9cb62cf692a0c64e7bf38058178c293d33496dd6",
+  "identical": true
+ },
+ "truthfulqa": {
+  "parent_rows": 4804,
+  "output_rows": 4804,
+  "sha256_parent": "31d03f6bd6a6cc35cebe15e4a6085c72e061b5629d9ba4e23daf4d2ad5b81ecd",
+  "sha256_output": "31d03f6bd6a6cc35cebe15e4a6085c72e061b5629d9ba4e23daf4d2ad5b81ecd",
+  "identical": true
+ },
+ "ultrachat": {
+  "parent_rows": 8000,
+  "output_rows": 8000,
+  "sha256_parent": "d2aa9238ec8951757a8e2c5b3f65fdd35bcaeed5f8be3828a6051effc24706ed",
+  "sha256_output": "d2aa9238ec8951757a8e2c5b3f65fdd35bcaeed5f8be3828a6051effc24706ed",
+  "identical": true
+ },
+ "werewolf": {
+  "parent_rows": 6692,
+  "output_rows": 6692,
+  "sha256_parent": "a18582dbb566c9d98bb3908251850c0f26b57040953c365773c500507c9ecaeb",
+  "sha256_output": "a18582dbb566c9d98bb3908251850c0f26b57040953c365773c500507c9ecaeb",
+  "identical": true
+ },
+ "wikihow": {
+  "parent_rows": 1760,
+  "output_rows": 1760,
+  "sha256_parent": "ea4b600ee8f5b009edd72dadf416a7349e69b0c07f7474b1c6e9621df8e5db98",
+  "sha256_output": "ea4b600ee8f5b009edd72dadf416a7349e69b0c07f7474b1c6e9621df8e5db98",
+  "identical": true
+ },
+ "wikihownest": {
+  "parent_rows": 7960,
+  "output_rows": 7960,
+  "sha256_parent": "d4cc5eb14642b5dde835bf7ce62426ff4f23b42472127f70031d4d56687dae5f",
+  "sha256_output": "d4cc5eb14642b5dde835bf7ce62426ff4f23b42472127f70031d4d56687dae5f",
+  "identical": true
+ }
+}