Datasets:

EricSpencer00
/

chattla-rft-corpora-v2

Error code:   DatasetGenerationError
Exception:    CastError
Message:      Couldn't cast
source: string
module: string
tier: string
distinct_states: int64
complexity_score: double
safety_catch_rate: double
quality_gold: bool
reward_weight: double
features: struct<num_variables: int64, num_definitions: int64, num_disjuncts: int64, num_conjuncts: int64, tem (... 67 chars omitted)
  child 0, num_variables: int64
  child 1, num_definitions: int64
  child 2, num_disjuncts: int64
  child 3, num_conjuncts: int64
  child 4, temporal_op_count: int64
  child 5, extends_depth: int64
  child 6, noncomment_loc: int64
quality_fail_reasons: list<item: string>
  child 0, item: string
safety_killed: int64
vacuity: list<item: string>
  child 0, item: string
temporal_op_count: int64
num_disjuncts: int64
kill_rate: double
num_definitions: int64
comment_ratio: double
num_variables: int64
num_conjuncts: int64
mutant_violations: list<item: string>
  child 0, item: string
noncomment_loc: int64
mutants_attempted: int64
extends_depth: int64
to
{'source': Value('string'), 'module': Value('string'), 'tier': Value('string'), 'complexity_score': Value('float64'), 'num_variables': Value('int64'), 'num_definitions': Value('int64'), 'extends_depth': Value('int64'), 'num_disjuncts': Value('int64'), 'num_conjuncts': Value('int64'), 'temporal_op_count': Value('int64'), 'noncomment_loc': Value('int64'), 'comment_ratio': Value('float64'), 'distinct_states': Value('int64'), 'vacuity': List(Value('string')), 'safety_catch_rate': Value('float64'), 'kill_rate': Value('float64'), 'mutants_attempted': Value('int64'), 'safety_killed': Value('int64'), 'mutant_violations': List(Value('string')), 'quality_gold': Value('bool'), 'quality_fail_reasons': List(Value('string'))}
because column names don't match
Traceback:    Traceback (most recent call last):
                File "/usr/local/lib/python3.14/site-packages/datasets/builder.py", line 1816, in _prepare_split_single
                  for key, table in generator:
                                    ^^^^^^^^^
                File "/src/services/worker/src/worker/job_runners/config/parquet_and_info.py", line 613, in wrapped
                  for item in generator(*args, **kwargs):
                              ~~~~~~~~~^^^^^^^^^^^^^^^^^
                File "/usr/local/lib/python3.14/site-packages/datasets/packaged_modules/json/json.py", line 343, in _generate_tables
                  self._cast_table(pa_table, json_field_paths=json_field_paths),
                  ~~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
                File "/usr/local/lib/python3.14/site-packages/datasets/packaged_modules/json/json.py", line 132, in _cast_table
                  pa_table = table_cast(pa_table, self.info.features.arrow_schema)
                File "/usr/local/lib/python3.14/site-packages/datasets/table.py", line 2369, in table_cast
                  return cast_table_to_schema(table, schema)
                File "/usr/local/lib/python3.14/site-packages/datasets/table.py", line 2297, in cast_table_to_schema
                  raise CastError(
                  ...<3 lines>...
                  )
              datasets.table.CastError: Couldn't cast
              source: string
              module: string
              tier: string
              distinct_states: int64
              complexity_score: double
              safety_catch_rate: double
              quality_gold: bool
              reward_weight: double
              features: struct<num_variables: int64, num_definitions: int64, num_disjuncts: int64, num_conjuncts: int64, tem (... 67 chars omitted)
                child 0, num_variables: int64
                child 1, num_definitions: int64
                child 2, num_disjuncts: int64
                child 3, num_conjuncts: int64
                child 4, temporal_op_count: int64
                child 5, extends_depth: int64
                child 6, noncomment_loc: int64
              quality_fail_reasons: list<item: string>
                child 0, item: string
              safety_killed: int64
              vacuity: list<item: string>
                child 0, item: string
              temporal_op_count: int64
              num_disjuncts: int64
              kill_rate: double
              num_definitions: int64
              comment_ratio: double
              num_variables: int64
              num_conjuncts: int64
              mutant_violations: list<item: string>
                child 0, item: string
              noncomment_loc: int64
              mutants_attempted: int64
              extends_depth: int64
              to
              {'source': Value('string'), 'module': Value('string'), 'tier': Value('string'), 'complexity_score': Value('float64'), 'num_variables': Value('int64'), 'num_definitions': Value('int64'), 'extends_depth': Value('int64'), 'num_disjuncts': Value('int64'), 'num_conjuncts': Value('int64'), 'temporal_op_count': Value('int64'), 'noncomment_loc': Value('int64'), 'comment_ratio': Value('float64'), 'distinct_states': Value('int64'), 'vacuity': List(Value('string')), 'safety_catch_rate': Value('float64'), 'kill_rate': Value('float64'), 'mutants_attempted': Value('int64'), 'safety_killed': Value('int64'), 'mutant_violations': List(Value('string')), 'quality_gold': Value('bool'), 'quality_fail_reasons': List(Value('string'))}
              because column names don't match
              
              The above exception was the direct cause of the following exception:
              
              Traceback (most recent call last):
                File "/src/services/worker/src/worker/job_runners/config/parquet_and_info.py", line 1369, in compute_config_parquet_and_info_response
                  parquet_operations, partial, estimated_dataset_info = stream_convert_to_parquet(
                                                                        ~~~~~~~~~~~~~~~~~~~~~~~~~^
                      builder, max_dataset_size_bytes=max_dataset_size_bytes
                      ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
                  )
                  ^
                File "/src/services/worker/src/worker/job_runners/config/parquet_and_info.py", line 948, in stream_convert_to_parquet
                  builder._prepare_split(split_generator=splits_generators[split], file_format="parquet")
                  ~~~~~~~~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
                File "/usr/local/lib/python3.14/site-packages/datasets/builder.py", line 1683, in _prepare_split
                  for job_id, done, content in self._prepare_split_single(
                                               ~~~~~~~~~~~~~~~~~~~~~~~~~~^
                      gen_kwargs=gen_kwargs, job_id=job_id, **_prepare_split_args
                      ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
                  ):
                  ^
                File "/usr/local/lib/python3.14/site-packages/datasets/builder.py", line 1869, in _prepare_split_single
                  raise DatasetGenerationError("An error occurred while generating the dataset") from e
              datasets.exceptions.DatasetGenerationError: An error occurred while generating the dataset

Need help to make the dataset viewer work? Make sure to review how to configure the dataset viewer, and open a discussion for direct support.

source string	module string	tier string	complexity_score float64	num_variables int64	num_definitions int64	extends_depth int64	num_disjuncts int64	num_conjuncts int64	temporal_op_count int64	noncomment_loc int64	comment_ratio float64	distinct_states int64	vacuity list	safety_catch_rate float64	kill_rate float64	mutants_attempted int64	safety_killed int64	mutant_violations list	quality_gold bool	quality_fail_reasons list
data/raw/1j01/mopaint/tla/Mopaint.tla	Mopaint	tier1_sany_cfg	184.2	10	17	3	10	104	2	192	0.068	25	[]	null	null	0	0	[]	false	[ "weak_mutation" ]
data/raw/Alexander-N/tla-specs/aiorwlock/aiorwlock.tla	aiorwlock	tier1_sany_cfg	92.1	1	15	4	10	46	2	56	0	17	[]	null	null	0	0	[]	false	[ "weak_mutation" ]
data/raw/Alexander-N/tla-specs/aiorwlock/aiorwlock_bug.tla	aiorwlock_bug	tier1_sany_cfg	80.2	1	15	4	10	35	2	47	0	18	[]	null	null	0	0	[]	false	[ "weak_mutation" ]
data/raw/Alexander-N/tla-specs/asyncio-lock/step1.tla	step1	tier1_sany_cfg	94.2	3	14	2	6	36	8	82	0.146	295	[]	null	null	0	0	[]	false	[ "weak_mutation" ]
data/raw/Alexander-N/tla-specs/asyncio-lock/step2.tla	step2	tier1_sany_cfg	46.2	3	9	2	5	9	4	37	0.075	116	[]	null	null	0	0	[]	false	[ "weak_mutation" ]
data/raw/Alexander-N/tla-specs/asyncio-lock/step3.tla	step3	tier1_sany_cfg	57.1	3	11	2	5	16	4	46	0.098	110	[]	null	null	0	0	[]	false	[ "weak_mutation" ]
data/raw/Alexander-N/tla-specs/asyncio-lock/step4.tla	step4	tier1_sany_cfg	47.8	3	10	2	5	9	4	38	0.073	116	[]	null	null	0	0	[]	false	[ "weak_mutation" ]
data/raw/Alexander-N/tla-specs/crdt-bug/crdt/crdt.tla	crdt	tier1_sany_cfg	81.8	3	18	4	7	23	5	68	0	138	[]	null	null	0	0	[]	false	[ "weak_mutation" ]
data/raw/Alexander-N/tla-specs/crdt-bug/crdt_fixed/crdt_fixed.tla	crdt_fixed	tier1_sany_cfg	81.1	3	18	4	6	23	5	71	0	566	[]	null	null	0	0	[]	false	[ "weak_mutation" ]
data/raw/Alexander-N/tla-specs/dual-writes/channels/ChannelsReliable.tla	ChannelsReliable	tier1_sany_cfg	19.3	0	8	4	0	0	2	13	0.235	null	[]	null	null	0	0	[]	false	[ "weak_mutation" ]
data/raw/Alexander-N/tla-specs/dual-writes/channels/ChannelsUtils.tla	ChannelsUtils	tier1_sany_cfg	9.7	0	3	2	0	0	0	42	0.222	null	[]	null	null	0	0	[]	false	[ "weak_mutation" ]
data/raw/Alexander-N/tla-specs/dual-writes/channels/dual_writes_channels.tla	dual_writes_channels	tier1_sany_cfg	125.2	5	20	3	6	52	7	117	0.071	49	[]	null	null	0	0	[]	false	[ "weak_mutation" ]
data/raw/Alexander-N/tla-specs/dual-writes/shiviz/dual_writes_shiviz.tla	dual_writes_shiviz	tier1_sany_cfg	134.2	5	20	3	8	55	9	117	0.017	79	[]	null	null	0	0	[]	false	[ "weak_mutation" ]
data/raw/Alexander-N/tla-specs/dual-writes/simple/dual_writes.tla	dual_writes	tier1_sany_cfg	54.6	2	13	2	2	14	5	41	0.047	15	[]	null	null	0	0	[]	false	[ "weak_mutation" ]
data/raw/Alexander-N/tla-specs/dual-writes/vector_clock/dual_writes_vector_clock.tla	dual_writes_vector_clock	tier1_sany_cfg	178.3	9	21	4	6	82	9	208	0.067	201	[]	null	null	0	0	[]	false	[ "weak_mutation" ]
data/raw/Alexander-N/tla-specs/paxos-from-the-ground-up/01_race.tla	01_race	tier1_sany_cfg	53.9	0	15	3	3	12	5	49	0.02	81	[]	null	null	0	0	[]	false	[ "weak_mutation" ]
data/raw/Alexander-N/tla-specs/paxos-from-the-ground-up/01_race_pluscal.tla	01_race_pluscal	tier1_sany_cfg	87.1	3	14	3	2	27	10	96	0.04	81	[]	null	null	0	0	[]	false	[ "weak_mutation" ]
data/raw/Alexander-N/tla-specs/paxos-from-the-ground-up/02_deadlock.tla	02_deadlock	tier1_sany_cfg	88.7	0	16	3	11	32	6	82	0.035	305	[]	null	null	0	0	[]	false	[ "weak_mutation" ]
data/raw/Alexander-N/tla-specs/paxos-from-the-ground-up/messaging.tla	messaging	tier1_sany_cfg	34.8	0	10	3	0	12	1	43	0	null	[]	null	null	0	0	[]	false	[ "weak_mutation" ]
data/raw/Alexander-N/tla-specs/paxos-from-the-ground-up/messagingPlusCal.tla	messagingPlusCal	tier1_sany_cfg	11.7	0	3	3	0	2	1	17	0	null	[]	null	null	0	0	[]	false	[ "weak_mutation" ]
data/raw/Alexander-N/tla-specs/rate-limiter/rate_limiter.tla	rate_limiter	tier1_sany_cfg	138.2	5	22	2	13	49	11	102	0.089	48	[]	null	null	0	0	[]	false	[ "weak_mutation" ]
data/raw/Alexander-N/tla-specs/tic-tac-toe/tic_tac_toe.tla	tic_tac_toe	tier1_sany_cfg	93.5	2	32	4	8	21	1	85	0.045	null	[]	null	null	0	0	[]	false	[ "weak_mutation" ]
data/raw/Alexander-N/tla-specs/tower-of-hanoi/MC.tla	MC	tier1_sany_cfg	9.6	0	3	1	0	0	2	6	0	27	[]	1	1	1	1	[ "TotalConstant" ]	true	[]
data/raw/AntonGames/SEMM-lab1/Lab1a/Kerberos.tla	Kerberos	tier1_sany_cfg	141.6	0	21	1	16	65	8	126	0.227	5,212	[]	0	0	1	0	[]	false	[ "weak_mutation" ]
data/raw/AntonGames/SEMM-lab1/Lab1a/KerberosLive.tla	KerberosLive	tier1_sany_cfg	0.8	0	0	1	0	0	0	3	0.25	246	[]	null	null	0	0	[]	false	[ "weak_mutation" ]
data/raw/AntonGames/SEMM-lab1/Lab1b/KerberosAbstract.tla	KerberosAbstract	tier1_sany_cfg	34.7	6	7	1	0	4	3	17	0.227	4	[]	null	null	0	0	[]	false	[ "weak_mutation" ]
data/raw/AntonGames/SEMM-lab1/Lab1b/KerberosRefinement.tla	KerberosRefinement	tier1_sany_cfg	127.6	0	19	2	10	61	5	171	0.095	246	[]	0	0.5	2	0	[ "TypeOK" ]	false	[ "weak_mutation" ]
data/raw/AntonGames/SEMM-lab1/lectures/Example.tla	Example	tier1_sany_cfg	14.3	1	3	0	0	3	2	8	0	2	[ "no_invariant_or_property_in_cfg" ]	null	null	0	0	[]	false	[ "vacuous", "thin_model", "weak_mutation" ]
data/raw/AntonGames/SEMM-lab1/lectures/MyGCD/MyGCD.tla	MyGCD	tier1_sany_cfg	58.6	2	12	1	2	18	6	41	0.068	4	[]	null	null	0	0	[]	false	[ "weak_mutation" ]
data/raw/AntonGames/SEMM-lab1/lectures/MyGCD/MyGCD_Check.tla	MyGCD_Check	tier1_sany_cfg	9.1	0	2	1	0	1	2	6	0.143	3	[]	0	0	1	0	[]	false	[ "weak_mutation" ]
data/raw/AntonGames/SEMM-lab1/lectures/MyGCD/MyGCD_Magic.tla	MyGCD_Magic	tier1_sany_cfg	43.1	2	10	1	0	13	4	26	0	4	[]	null	null	0	0	[]	false	[ "weak_mutation" ]
data/raw/AntonGames/SEMM-lab1/lectures/Rounds/Round.tla	Round	tier1_sany_cfg	54.1	3	11	1	1	20	3	41	0.068	60	[]	null	null	0	0	[]	false	[ "weak_mutation" ]
data/raw/CAPHTECH/kiri/docs/formal/PathPenaltyEncoding.tla	PathPenaltyEncoding	tier1_sany_cfg	21.1	0	9	2	0	2	0	46	0	null	[]	null	null	0	0	[]	false	[ "weak_mutation" ]
data/raw/CAPHTECH/kiri/docs/formal/PathPenaltyMerge.tla	PathPenaltyMerge	tier1_sany_cfg	97.3	3	34	4	0	12	4	183	0	193	[]	0.5	0.5	2	1	[ "EncodeDecodeInvariant" ]	true	[]
data/raw/CAPHTECH/kiri/docs/formal/language-support/PlanA_CentralRegistry.tla	PlanA_CentralRegistry	tier1_sany_cfg	61.6	0	14	3	3	26	1	81	0.047	0	[]	null	null	0	0	[]	false	[ "thin_model", "weak_mutation" ]
data/raw/CAPHTECH/kiri/docs/formal/language-support/PlanB_Pool.tla	PlanB_Pool	tier1_sany_cfg	78.5	0	17	4	4	35	1	100	0.01	0	[]	null	null	0	0	[]	false	[ "thin_model", "weak_mutation" ]
data/raw/CAPHTECH/kiri/docs/formal/language-support/PlanC_Capability.tla	PlanC_Capability	tier1_sany_cfg	86.2	0	19	3	4	39	1	112	0.009	0	[]	null	null	0	0	[]	false	[ "thin_model", "weak_mutation" ]
data/raw/CPChain/chain/tlaplus/Consensus.tla	Consensus	tier1_sany_cfg	140.9	1	9	3	3	36	28	289	0.077	null	[]	null	null	0	0	[]	false	[ "weak_mutation" ]
data/raw/CPChain/chain/tlaplus/lbft.tla	lbft	tier1_sany_cfg	4,010.4	4	325	3	631	2,606	2	2,724	0.012	null	[]	null	null	0	0	[]	false	[ "weak_mutation" ]
data/raw/CPChain/chain/tlaplus/lbft2.tla	lbft2	tier1_sany_cfg	250.9	11	30	3	25	123	2	304	0.208	null	[]	null	null	0	0	[]	false	[ "weak_mutation" ]
data/raw/CPChain/chain/tlaplus/lbft3.tla	lbft3	tier1_sany_cfg	374.9	10	40	4	30	192	7	569	0.1	332	[]	null	null	0	0	[]	false	[ "weak_mutation" ]
data/raw/CPChain/chain/tlaplus/lbft4.tla	lbft4	tier1_sany_cfg	366.3	11	39	4	30	187	7	528	0.137	884	[]	null	null	0	0	[]	false	[ "weak_mutation" ]
data/raw/Calvin-L/ezpsl/examples/Assertions.tla	Assertions	tier1_sany_cfg	188.1	6	14	3	12	115	6	146	0.02	17	[]	null	null	0	0	[]	false	[ "weak_mutation" ]
data/raw/Calvin-L/ezpsl/examples/If.tla	If	tier1_sany_cfg	146.6	6	12	3	9	85	5	111	0.026	4	[]	null	null	0	0	[]	false	[ "weak_mutation" ]
data/raw/Calvin-L/ezpsl/examples/NondeterministicAssignment.tla	NondeterministicAssignment	tier1_sany_cfg	137.3	6	12	4	8	77	5	103	0.028	24	[]	null	null	0	0	[]	false	[ "weak_mutation" ]
data/raw/Calvin-L/ezpsl/examples/Refinement.tla	Refinement	tier1_sany_cfg	203.4	7	17	3	11	120	8	154	0.019	58	[]	null	null	0	0	[]	false	[ "weak_mutation" ]
data/raw/Calvin-L/ezpsl/examples/RefinementAbstract.tla	RefinementAbstract	tier1_sany_cfg	18.7	2	4	1	0	5	1	12	0	null	[]	null	null	0	0	[]	false	[ "weak_mutation" ]
data/raw/Calvin-L/ezpsl/examples/Stack.tla	Stack	tier1_sany_cfg	590.5	6	40	3	38	415	7	500	0.006	1,606	[ "no_invariant_or_property_in_cfg" ]	null	null	0	0	[]	false	[ "vacuous", "weak_mutation" ]
data/raw/Calvin-L/ezpsl/examples/Test.tla	Test	tier1_sany_cfg	154.9	6	13	3	9	91	5	119	0.025	20	[]	null	null	0	0	[]	false	[ "weak_mutation" ]
data/raw/Calvin-L/ezpsl/examples/While.tla	While	tier1_sany_cfg	142.2	6	12	3	9	81	5	107	0.027	5	[]	null	null	0	0	[]	false	[ "weak_mutation" ]
data/raw/CharlesHoskinson/ProjectLegends/spec/tla/APIContract.tla	APIContract	tier1_sany_cfg	302	0	50	4	14	147	14	360	0.161	277	[]	0	0.5	2	0	[ "TypeOK" ]	false	[ "weak_mutation" ]
data/raw/CharlesHoskinson/ProjectLegends/spec/tla/BusMinimal.tla	BusMinimal	tier1_sany_cfg	82.9	2	23	4	0	25	5	74	0.267	1	[ "only_1_distinct_states" ]	null	null	0	0	[]	false	[ "vacuous", "thin_model", "weak_mutation" ]
data/raw/CharlesHoskinson/ProjectLegends/spec/tla/Capture.tla	Capture	tier1_sany_cfg	161.6	0	25	4	10	56	17	221	0.068	732	[]	0	1	1	0	[ "TypeOK" ]	false	[ "weak_mutation" ]
data/raw/CharlesHoskinson/ProjectLegends/spec/tla/CaptureMinimal.tla	CaptureMinimal	tier1_sany_cfg	110.2	0	20	2	8	40	9	132	0.07	36	[]	null	null	0	0	[]	false	[ "weak_mutation" ]
data/raw/CharlesHoskinson/ProjectLegends/spec/tla/ConfigValidation.tla	ConfigValidation	tier1_sany_cfg	87.3	0	15	2	4	42	1	158	0.048	595	[]	null	null	0	0	[]	false	[ "weak_mutation" ]
data/raw/CharlesHoskinson/ProjectLegends/spec/tla/DMA.tla	DMA	tier1_sany_cfg	67.5	0	22	4	1	13	0	185	0.119	null	[]	null	null	0	0	[]	false	[ "weak_mutation" ]
data/raw/CharlesHoskinson/ProjectLegends/spec/tla/DMATest.tla	DMATest	tier1_sany_cfg	100.2	0	25	4	5	37	3	127	0.287	73	[]	null	null	0	0	[]	false	[ "weak_mutation" ]
data/raw/CharlesHoskinson/ProjectLegends/spec/tla/Determinism.tla	Determinism	tier1_sany_cfg	190.6	0	24	4	9	69	23	286	0.04	960	[]	null	null	0	0	[]	false	[ "weak_mutation" ]
data/raw/CharlesHoskinson/ProjectLegends/spec/tla/EmuKernel.tla	EmuKernel	tier1_sany_cfg	201.1	0	45	6	8	75	10	276	0.207	null	[]	null	null	0	0	[]	false	[ "weak_mutation" ]
data/raw/CharlesHoskinson/ProjectLegends/spec/tla/ErrorModel.tla	ErrorModel	tier1_sany_cfg	96.7	0	21	2	2	38	1	222	0.083	2,017	[]	0	1	1	0	[ "TypeOK" ]	false	[ "weak_mutation" ]
data/raw/CharlesHoskinson/ProjectLegends/spec/tla/Input.tla	Input	tier1_sany_cfg	155.4	0	26	4	7	70	5	274	0.096	743	[]	null	null	0	0	[]	false	[ "weak_mutation" ]
data/raw/CharlesHoskinson/ProjectLegends/spec/tla/InputMinimal.tla	InputMinimal	tier1_sany_cfg	105.2	0	17	4	6	53	3	127	0.093	4,717	[]	0	0	2	0	[]	false	[ "weak_mutation" ]
data/raw/CharlesHoskinson/ProjectLegends/spec/tla/InterruptTest.tla	InterruptTest	tier1_sany_cfg	106.8	0	33	4	3	30	6	103	0.309	7	[]	null	null	0	0	[]	false	[ "weak_mutation" ]
data/raw/CharlesHoskinson/ProjectLegends/spec/tla/Lifecycle.tla	Lifecycle	tier1_sany_cfg	210.1	0	26	4	8	113	7	341	0.02	235,160	[]	0	1	1	0	[ "TypeOK" ]	false	[ "weak_mutation" ]
data/raw/CharlesHoskinson/ProjectLegends/spec/tla/LifecycleMinimal.tla	LifecycleMinimal	tier1_sany_cfg	155.5	0	23	2	11	89	1	180	0.091	198	[]	0	1	2	0	[ "TypeOK", "TypeOK" ]	false	[ "weak_mutation" ]
data/raw/CharlesHoskinson/ProjectLegends/spec/tla/PAL.tla	PAL	tier1_sany_cfg	205	0	33	4	11	84	14	305	0.053	547	[]	0	1	1	0	[ "TypeOK" ]	false	[ "weak_mutation" ]
data/raw/CharlesHoskinson/ProjectLegends/spec/tla/PALMinimal.tla	PALMinimal	tier1_sany_cfg	107.7	0	21	2	8	52	1	132	0.102	3,533	[]	0	0	1	0	[]	false	[ "weak_mutation" ]
data/raw/CharlesHoskinson/ProjectLegends/spec/tla/PIC.tla	PIC	tier1_sany_cfg	64.8	0	26	4	0	6	0	178	0.168	null	[]	null	null	0	0	[]	false	[ "weak_mutation" ]
data/raw/CharlesHoskinson/ProjectLegends/spec/tla/PIT.tla	PIT	tier1_sany_cfg	58.5	0	19	4	0	6	3	160	0.144	null	[]	null	null	0	0	[]	false	[ "weak_mutation" ]
data/raw/CharlesHoskinson/ProjectLegends/spec/tla/Reentrancy.tla	Reentrancy	tier1_sany_cfg	121.2	0	20	2	9	59	2	182	0.057	null	[]	null	null	0	0	[]	false	[ "weak_mutation" ]
data/raw/CharlesHoskinson/ProjectLegends/spec/tla/ReentrancyMinimal.tla	ReentrancyMinimal	tier1_sany_cfg	99.5	0	17	2	8	51	1	120	0.04	34	[]	0	1	1	0	[ "TypeOK" ]	false	[ "weak_mutation" ]
data/raw/CharlesHoskinson/ProjectLegends/spec/tla/SaveState.tla	SaveState	tier1_sany_cfg	206.6	0	57	4	5	63	10	311	0.139	2	[]	null	null	0	0	[]	false	[ "thin_model", "weak_mutation" ]
data/raw/CharlesHoskinson/ProjectLegends/spec/tla/SaveStateTest.tla	SaveStateTest	tier1_sany_cfg	168.2	0	41	4	7	63	4	267	0.116	5	[]	0	0	1	0	[]	false	[ "weak_mutation" ]
data/raw/CharlesHoskinson/ProjectLegends/spec/tla/Scheduler.tla	Scheduler	tier1_sany_cfg	85.5	0	20	5	0	8	9	270	0.135	null	[]	null	null	0	0	[]	false	[ "weak_mutation" ]
data/raw/CharlesHoskinson/ProjectLegends/spec/tla/SchedulerMinimal.tla	SchedulerMinimal	tier1_sany_cfg	104.8	0	23	4	4	37	9	93	0.295	14	[]	null	null	0	0	[]	false	[ "weak_mutation" ]
data/raw/CharlesHoskinson/ProjectLegends/spec/tla/SchedulerTest.tla	SchedulerTest	tier1_sany_cfg	202.2	0	31	6	16	74	21	207	0.191	493	[]	0	0.5	2	0	[ "TypeOK" ]	false	[ "weak_mutation" ]
data/raw/CharlesHoskinson/ProjectLegends/spec/tla/Threading.tla	Threading	tier1_sany_cfg	220.4	0	34	4	17	105	8	294	0.067	82	[]	0	0	1	0	[]	false	[ "weak_mutation" ]
data/raw/CharlesHoskinson/ProjectLegends/spec/tla/ThreadingMinimal.tla	ThreadingMinimal	tier1_sany_cfg	178.9	0	27	3	15	96	3	199	0.095	1,072	[]	0	0	1	0	[]	false	[ "weak_mutation" ]
data/raw/CharlesHoskinson/ProjectLegends/spec/tla/Types.tla	Types	tier1_sany_cfg	147.7	0	45	3	0	7	21	297	0.149	null	[]	null	null	0	0	[]	false	[ "weak_mutation" ]
data/raw/CharlesHoskinson/jolt-tla/JoltContinuations.tla	JoltContinuations	tier1_sany_cfg	436.6	1	169	4	5	73	19	631	0.102	null	[]	null	null	0	0	[]	false	[ "weak_mutation" ]
data/raw/Cjen1/consensus-tlaplus/LogPaxos/LogPaxos.tla	LogPaxos	tier1_sany_cfg	119.1	1	23	4	8	51	4	136	0	null	[]	null	null	0	0	[]	false	[ "weak_mutation" ]
data/raw/Cjen1/consensus-tlaplus/Paxos/Paxos.tla	Paxos	tier1_sany_cfg	105.6	1	21	4	8	48	1	121	0	null	[]	null	null	0	0	[]	false	[ "weak_mutation" ]
data/raw/Cjen1/consensus-tlaplus/conspire/Abstract.tla	Abstract	tier1_sany_cfg	68.7	0	12	2	12	28	1	77	0.252	null	[]	null	null	0	0	[]	false	[ "weak_mutation" ]
data/raw/Cjen1/consensus-tlaplus/conspire/AbstractLog.tla	AbstractLog	tier1_sany_cfg	78.7	0	15	3	8	34	2	87	0.25	null	[]	null	null	0	0	[]	false	[ "weak_mutation" ]
data/raw/Cjen1/consensus-tlaplus/conspire/PNoC.tla	PNoC	tier1_sany_cfg	56.3	0	19	4	2	15	1	68	0.105	20	[]	0	1	1	0	[ null ]	false	[ "weak_mutation" ]
data/raw/Cjen1/consensus-tlaplus/msfp/Nezha.tla	Nezha	tier1_sany_cfg	41.6	2	7	3	2	18	0	56	0.051	null	[]	null	null	0	0	[]	false	[ "weak_mutation" ]
data/raw/Cjen1/consensus-tlaplus/msfp/PrimedValue.tla	PrimedValue	tier1_sany_cfg	3.1	0	1	2	0	0	0	6	0	null	[]	null	null	0	0	[]	false	[ "weak_mutation" ]
data/raw/Cjen1/consensus-tlaplus/msfp/SimpleValue.tla	SimpleValue	tier1_sany_cfg	45.8	0	11	3	3	18	1	48	0	16,988	[]	0	0	1	0	[]	false	[ "weak_mutation" ]
data/raw/Cjen1/consensus-tlaplus/u2pc/U2PC.tla	U2PC	tier1_sany_cfg	167.3	0	27	4	16	91	1	158	0.233	null	[]	null	null	0	0	[]	false	[ "weak_mutation" ]
data/raw/Cjen1/consensus-tlaplus/u2pc/U2PC_MC.tla	U2PC_MC	tier1_sany_cfg	44.5	0	14	3	5	11	0	60	0.333	339	[]	0	1	1	0	[ null ]	false	[ "weak_mutation" ]
data/raw/DistCompiler/pgo/antithesis/wiredtiger/Storage.tla	Storage	tier1_sany_cfg	296.2	7	55	5	25	141	2	272	0.282	null	[]	null	null	0	0	[]	false	[ "weak_mutation" ]
data/raw/DistCompiler/pgo/omnilink/atomic/Atomic.tla	Atomic	tier1_sany_cfg	10.1	1	3	1	0	2	0	11	0	null	[]	null	null	0	0	[]	false	[ "weak_mutation" ]
data/raw/DistCompiler/pgo/omnilink/concurrentqueue/ConcurrentQueue.tla	ConcurrentQueue	tier1_sany_cfg	117.1	4	16	6	10	41	11	91	0.032	null	[]	null	null	0	0	[]	false	[ "weak_mutation" ]
data/raw/DistCompiler/pgo/omnilink/concurrentqueue/MCConcurrentQueue.tla	MCConcurrentQueue	tier1_sany_cfg	33.2	1	9	2	0	6	4	27	0	973	[]	null	null	0	0	[]	false	[ "weak_mutation" ]
data/raw/DistCompiler/pgo/omnilink/setbench/SetBench.tla	SetBench	tier1_sany_cfg	45	1	10	2	6	14	1	50	0	64	[]	0	0	1	0	[]	false	[ "weak_mutation" ]
data/raw/DistCompiler/pgo/omnilink/wiredtiger/MCStorage2Verso.tla	MCStorage2Verso	tier1_sany_cfg	47.9	1	10	3	0	14	4	74	0.129	null	[]	null	null	0	0	[]	false	[ "weak_mutation" ]
data/raw/DistCompiler/pgo/omnilink/wiredtiger/RefLocking.tla	RefLocking	tier1_sany_cfg	25	1	6	1	4	5	1	25	0	null	[]	null	null	0	0	[]	false	[ "weak_mutation" ]
data/raw/DistCompiler/pgo/omnilink/wiredtiger/Storage.tla	Storage	tier1_sany_cfg	296.7	7	54	5	27	141	2	272	0.282	null	[]	null	null	0	0	[]	false	[ "weak_mutation" ]
data/raw/DistCompiler/pgo/omnilink/wiredtiger/Storage2.tla	Storage2	tier1_sany_cfg	312	7	53	6	29	152	3	285	0.267	null	[]	null	null	0	0	[]	false	[ "weak_mutation" ]
data/raw/DistCompiler/pgo/systems/dqueue/dqueue.tla	dqueue	tier1_sany_cfg	128.2	5	17	3	6	50	8	192	0.05	635	[]	0	0	1	0	[]	false	[ "weak_mutation" ]

End of preview.

ChatTLA RFT Corpora v2 — verifier-gated TLA+ generation corpus

Rejection-sampling fine-tuning (RFT/STaR) corpus for TLA+ specification generation, produced by the prove-TLA verify-until-correct loop (2026-07-10/11). Every survivor passed the full hard-metric gate chain — no LLM-judge scoring anywhere:

SANY parse → semantic-invariant cfg gate → TLC model-check (non-vacuous, ≥3 distinct states) → NL↔invariant linkage contract (PROPERTY_INVARIANT named, defined, checked, non-TypeOK) → liveness-checked-or-absent gate → deterministic mutation battery (crash-kill discounted, TypeOK-only catches rejected) → near-dup decontamination (Jaccard < 0.65 vs the 206-spec FormaLLM corpus, the frozen holdout-30, and tlaplus/examples).

Files

file	rows	what
`data/w2_survivors.jsonl`	111	Generated (NL, TLA+ module, TLC cfg) triples that survived the loop. Fields: `seed_key`, `nl` (natural-language spec w/ SAFETY PROPERTY section), `spec_text`, `cfg_text`, `property_invariant`, `distinct_states`, `vacuity`, `safety_catch_rate`, `mutation_evidence`, structural features, `complexity_score`, `reward_weight`, `decontam` similarity, `iters`.
`data/manifest_w2_seeds.jsonl`	120	v2 seed pool (non-vacuous, ≥3-state scraped specs) with adequacy soft-labels; generation seeds are back-translations anchored to these.
`data/manifest_w2_seeds_wide.jsonl`	501	v3-wide seed pool from the 242-repo re-scrape (4.2× v2), same criteria + labels.
`data/adequacy_labels_v3wide.jsonl`	1,229	Full adequacy-battery soft-labels over the wide corpus (TLC-derived `distinct_states`, vacuity flags, mutation `safety_catch_rate`/`kill_rate` with violated-invariant attribution, structural-complexity features, `quality_gold`).

Seed manifests reference scraped source paths + metrics only; third-party .tla file contents are not redistributed here (licensing) — sources are public GitHub repos recorded in each row's source field.

Generation provenance (hard numbers)

Sampler: gpt-oss-120b (stock, ALCF Sophia vLLM), temp 0.8, ≤8 verify-repair iterations per attempt. Yield by pass: pass-1 39/120 (32.5%), pass-2-on-rejects 10/81 (12.3%), k2-on-proven-seeds 28/49 (57.1%), k3 34/49 (69.4%). Cumulative 111/299 = 37.1%. Same-seed diversity: 1/83 near-dup pairs (median Jaccard 0.198). Survivor complexity: median 45.5 (max 180.5); distinct states median ~19 (max 139,860).

Known limitations (disclosed)

111 survivors derive from 49 unique seed specs — family concentration is real; family-stratified evaluation is required downstream.
69% of survivors have mutation_evidence="no_site" (deterministic battery found no applicable mutation site) — their invariant strength is unproven, not disproven.
0/111 check a liveness PROPERTY (models drop fairness rather than check it).
The NL↔invariant contract pins linkage structurally; it does not verify deep semantic equivalence of prose and invariant.

Downstream result (2026-07-12)

Fine-tuning gpt-oss-20b on this corpus (harmony final-channel rendering, experts-reaching LoRA) produced chattla-v2-sft2 — pass@4 = 2/30 on a frozen TLC-verified holdout, on specs the base model never passes at pass@32. First empirical evidence that verifier-gated RFT on this corpus transfers capability.

Part of the prove-TLA / ChatTLA project. Prior artifacts: chattla-tla-prover-corpora-v1, chattla-tla-prover-108-108.

W4 cross-family snapshot (added 2026-07-19)

data/w4_survivors_snapshot_20260719.jsonl — 1,690 rows. A point-in-time aggregate of every survivor row produced by the prove-TLA verify-until-correct loop as of 2026-07-19, dominated by the W4 cross-family generation (a 20 domain × 10 mechanism × 7 property × 6 twist scenario lattice). Same schema as data/w2_survivors.jsonl, same unchanged hard-metric gate chain (SANY → non-vacuous TLC → mutation battery → NL↔invariant fidelity → near-dup decontamination), no LLM-judge anywhere.

Cross-family arm comparison (per the project's weekly ledger): Claude-Opus teacher subagents cleared the lattice at ~100% cell yield; the gpt-oss-120b serve funnel at ~12.5%.

This is a partial, live snapshot — the generation loop was still running when this was captured (target floor ~5k rows before any fine-tune decision). Treat row counts and family balance as provisional; a frozen release will supersede this file.

Downloads last month: 63

Total file size:

8.41 MB

Models trained or fine-tuned on EricSpencer00/chattla-rft-corpora-v2

EricSpencer00/chattla-v2-sft2

Text Generation • 21B • Updated 25 days ago • 395