fork_turn -> fork_assistant_turn (1-based, assistant turns) + fork_turn_index (0-based, in turns)

README.md

@@ -77,7 +77,7 @@ Each conformant dialogue is forked at **four** evenly spaced points (*k* = 1, 2,
 <p align="center"><img src="https://huggingface.co/datasets/DialFork/dialfork/resolve/main/assets/forking.png" width="60%" alt="A conformant dialogue forked at four positions: the earlier the fork, the less similar to the original"></p>
 
 - **`fork_position`** (*k*): an earlier fork leaves less of the conversation conformant, so *k* = 1 > 2 > 3 > 4 is a built-in ranking of deviation;
-- **`fork_turn`** (*t<sub>k</sub>*): the assistant turn where the deviation starts.
+- **`fork_assistant_turn`** (*t<sub>k</sub>*): the assistant turn where the deviation starts, counting assistant turns only; `fork_turn_index` gives the same turn as a position in `turns`.
 
 ## 🧭 Four ways to go off-track
 
@@ -145,7 +145,7 @@ The labels come from the construction, so we checked that people see the same th
 
 - **Fork order is perceived.** Blind ratings rise with *k*: Kendall τ<sub>b</sub> = 0.64 within annotators (WIC, SSPC, GR).
 - **The injected behaviour is there.** Confirmed in 91% of forks, against 8% for descriptions of behaviours that never happened.
-- **It starts where intended.** 77% of the annotated onsets are within one turn of `fork_turn`.
+- **It starts where intended.** 77% of the annotated onsets are within one turn of `fork_assistant_turn`.
 - **The replayed beginning is clean.** 95% of the parts before the fork follow the workflow.
 
 The anonymized judgments and the annotation tool are released [on GitHub](https://github.com/idiap/dialfork/blob/main/docs/human_study.md).
@@ -158,7 +158,7 @@ from datasets import load_dataset
 ds = load_dataset("DialFork/dialfork", split="test")       # 36,278 rows, one per dialogue
 
 row = ds.filter(lambda r: r["id"] == "2115-GR-k2")[0]
-print(row["hallucination_type"], row["fork_position"], row["fork_turn"])   # goal_replacement 2 2
+print(row["hallucination_type"], row["fork_position"], row["fork_assistant_turn"])   # goal_replacement 2 2
 for turn in row["turns"][:4]:
     print(f'{turn["speaker"]:>6}: {turn["text"]}')
 ```
@@ -188,12 +188,13 @@ python scripts/download_data.py      # -> datasets/dialfork/ (+ datasets/STAR/)
 | `hallucination_type`, `hallucination_type_short` | string | e.g. `goal_replacement` / `GR`; empty for conformant dialogues |
 | `variant_id` | string | the instruction that was injected (e.g. `goal_collect_feedback`) |
 | `fork_position` | int | *k* in 1..4 (1 = earliest fork, most deviant); empty for conformant dialogues |
-| `fork_turn` | int | *t<sub>k</sub>*: the 1-based **assistant** turn the deviation starts at (counting assistant turns only) |
+| `fork_assistant_turn` | int | *t<sub>k</sub>*: the 1-based **assistant** turn the deviation starts at (counting assistant turns only); what Task 3 predicts |
+| `fork_turn_index` | int | the same turn as a 0-based position in `turns` (both speakers): `turns[:fork_turn_index]` is the part before the fork |
 | `num_turns` | int | number of turns (both speakers) |
 | `turns` | list | the conversation: `[{speaker, text}, ...]` |
 | `dialog_json` | string | the complete original file: personas, generation settings and seed, the full event trace (injected instruction included), and the task's workflow |
 
-A single `test` split: DialFork is an evaluation benchmark. Methods may use the conformant dialogues as reference data, but should never train on or tune with the labels (`fork_position`, `fork_turn`, `hallucination_type`).
+A single `test` split: DialFork is an evaluation benchmark. Methods may use the conformant dialogues as reference data, but should never train on or tune with the labels (`fork_position`, `fork_assistant_turn`, `fork_turn_index`, `hallucination_type`).
 
 ## 🎯 The benchmark
 

data/test-00000-of-00008.parquet

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:2247bf8415a6cdac7e6ec78f4f29746aac7ba82ee61f2a01217f210c4778a7d2
-size 4167508
+oid sha256:c828711d8418bcb3ec5fb816d33ecf6bb901d5f468b2396a067220b0939c9b17
+size 4170331

data/test-00001-of-00008.parquet

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:7ad893ec2921ef1a4dacdddb64be2bdad2bcfc59030bc07da11523628bc09cd8
-size 3549933
+oid sha256:f5b88dbcb050bdda3d03232a8cc4de9366f8ed0d0baaaffded8eb63625a49489
+size 3558178

data/test-00002-of-00008.parquet

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:15a6136c32f3df3ce364a337ef74cf4e15792553967cfe42f675d0bf3bed9232
-size 3218976
+oid sha256:0d8108ff61e9257ffec8258ae21bba281ececfad4610b2a825d912997846581e
+size 3228845

data/test-00003-of-00008.parquet

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:bf4c7186b7016a82bcd8afd24cd59b568937409ee7bfc7f50aac7082b82d3fdf
-size 4388683
+oid sha256:a5e82512b9d95a317000a32efd90a77c6518454c766698961d8277f549c32b99
+size 4396416

data/test-00004-of-00008.parquet

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:90e6bfd313317e4f073e8fdc3f68dcb5b45453bcbbc22f81a8a6271b35a6c957
-size 3477671
+oid sha256:f6172cecdbf9e2790d44b45deec6c4342790710cd4f6822856d7d257e62305fa
+size 3481806

data/test-00005-of-00008.parquet

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:94b14110fb411d3a54c5f908e25f79c67acb514fbf49a0f4303a401d702ccf61
-size 3679676
+oid sha256:f576535a66914fe7ad59f5d61c95157d92ed705ecd895993c7a83473002188e6
+size 3689840

data/test-00006-of-00008.parquet

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:b87ffaad242c92c38f656528c3323bbe4ae93457dfd3c6a20656d3c339b18439
-size 5240364
+oid sha256:30c05fbdfda06668b8c22809e040a8060710f315483b7bd47a17c630d11418ea
+size 5253271

data/test-00007-of-00008.parquet

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:b0f4585db882c1c27c2ed2242c6448ce3e016a6a1df95ad7e66b13af141400d6
-size 4662263
+oid sha256:0e77c0a4ea3a99fd5dad05293f5732c9c43ad574591fcfeb617c8554115bb970
+size 4672794