Datasets:

DebdipCS
/

cloze-congruence-multilingual-benchmark

Error code:   DatasetGenerationCastError
Exception:    DatasetGenerationCastError
Message:      An error occurred while generating the dataset

All the data files must have the same columns, but at some point there are 6 new columns ({'human_ratio', 'mix_regime', 'ai_ratio', 'sha256', 'paragraph_labels', 'word_count'}) and 6 missing columns ({'is_human', 'author', 'approx_word_count', 'sha256_hash', 'language_code', 'title'}).

This happened while the json dataset builder was generating data using

hf://datasets/DebdipCS/cloze-congruence-multilingual-benchmark/data/multilingual_mixed_human_ai_benchmark.json (at revision 1963369442b928953a3a0838618c098aaad34d28), ['hf://datasets/DebdipCS/cloze-congruence-multilingual-benchmark@1963369442b928953a3a0838618c098aaad34d28/data/multilingual_longform_benchmark.json', 'hf://datasets/DebdipCS/cloze-congruence-multilingual-benchmark@1963369442b928953a3a0838618c098aaad34d28/data/multilingual_mixed_human_ai_benchmark.json', 'hf://datasets/DebdipCS/cloze-congruence-multilingual-benchmark@1963369442b928953a3a0838618c098aaad34d28/data/real_world_downloaded_human_benchmark.json'], ['hf://datasets/DebdipCS/cloze-congruence-multilingual-benchmark@1963369442b928953a3a0838618c098aaad34d28/data/multilingual_longform_benchmark.json', 'hf://datasets/DebdipCS/cloze-congruence-multilingual-benchmark@1963369442b928953a3a0838618c098aaad34d28/data/multilingual_mixed_human_ai_benchmark.json', 'hf://datasets/DebdipCS/cloze-congruence-multilingual-benchmark@1963369442b928953a3a0838618c098aaad34d28/data/real_world_downloaded_human_benchmark.json']

Please either edit the data files to have matching columns, or separate them into different configurations (see docs at https://hf.co/docs/hub/datasets-manual-configuration#multiple-configurations)
Traceback:    Traceback (most recent call last):
                File "/usr/local/lib/python3.14/site-packages/datasets/builder.py", line 1848, in _prepare_split_single
                  writer.write_table(table)
                  ~~~~~~~~~~~~~~~~~~^^^^^^^
                File "/usr/local/lib/python3.14/site-packages/datasets/arrow_writer.py", line 765, in write_table
                  self._write_table(pa_table, writer_batch_size=writer_batch_size)
                  ~~~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
                File "/usr/local/lib/python3.14/site-packages/datasets/arrow_writer.py", line 773, in _write_table
                  pa_table = table_cast(pa_table, self._schema)
                File "/usr/local/lib/python3.14/site-packages/datasets/table.py", line 2378, in table_cast
                  return cast_table_to_schema(table, schema)
                File "/usr/local/lib/python3.14/site-packages/datasets/table.py", line 2306, in cast_table_to_schema
                  raise CastError(
                  ...<3 lines>...
                  )
              datasets.table.CastError: Couldn't cast
              doc_id: string
              language: string
              script: string
              mix_regime: string
              human_ratio: double
              ai_ratio: double
              paragraph_count: int64
              word_count: int64
              char_count: int64
              sha256: string
              paragraph_labels: list<item: string>
                child 0, item: string
              text: string
              to
              {'doc_id': Value('string'), 'language': Value('string'), 'language_code': Value('string'), 'script': Value('string'), 'author': Value('string'), 'is_human': Value('bool'), 'paragraph_count': Value('int64'), 'approx_word_count': Value('int64'), 'char_count': Value('int64'), 'sha256_hash': Value('string'), 'title': Value('string'), 'text': Value('string')}
              because column names don't match
              
              During handling of the above exception, another exception occurred:
              
              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 1694, 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 1850, in _prepare_split_single
                  raise DatasetGenerationCastError.from_cast_error(
                  ...<4 lines>...
                  )
              datasets.exceptions.DatasetGenerationCastError: An error occurred while generating the dataset
              
              All the data files must have the same columns, but at some point there are 6 new columns ({'human_ratio', 'mix_regime', 'ai_ratio', 'sha256', 'paragraph_labels', 'word_count'}) and 6 missing columns ({'is_human', 'author', 'approx_word_count', 'sha256_hash', 'language_code', 'title'}).
              
              This happened while the json dataset builder was generating data using
              
              hf://datasets/DebdipCS/cloze-congruence-multilingual-benchmark/data/multilingual_mixed_human_ai_benchmark.json (at revision 1963369442b928953a3a0838618c098aaad34d28), ['hf://datasets/DebdipCS/cloze-congruence-multilingual-benchmark@1963369442b928953a3a0838618c098aaad34d28/data/multilingual_longform_benchmark.json', 'hf://datasets/DebdipCS/cloze-congruence-multilingual-benchmark@1963369442b928953a3a0838618c098aaad34d28/data/multilingual_mixed_human_ai_benchmark.json', 'hf://datasets/DebdipCS/cloze-congruence-multilingual-benchmark@1963369442b928953a3a0838618c098aaad34d28/data/real_world_downloaded_human_benchmark.json'], ['hf://datasets/DebdipCS/cloze-congruence-multilingual-benchmark@1963369442b928953a3a0838618c098aaad34d28/data/multilingual_longform_benchmark.json', 'hf://datasets/DebdipCS/cloze-congruence-multilingual-benchmark@1963369442b928953a3a0838618c098aaad34d28/data/multilingual_mixed_human_ai_benchmark.json', 'hf://datasets/DebdipCS/cloze-congruence-multilingual-benchmark@1963369442b928953a3a0838618c098aaad34d28/data/real_world_downloaded_human_benchmark.json']
              
              Please either edit the data files to have matching columns, or separate them into different configurations (see docs at https://hf.co/docs/hub/datasets-manual-configuration#multiple-configurations)

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.

doc_id string	language string	language_code string	script string	author string	is_human bool	paragraph_count int64	approx_word_count int64	char_count int64	sha256_hash string	title string	text string
ML_DOC_001	English	en	Latin	Human_Academic	true	42	2,894	25,799	9353f1447079e67addd1cbb4346a2b256f371bf194213a482e4fe2c398c47f58	Long-Form Scholarly Treatise on Epistemic Systems (English - Human_Academic)	Proposition 1. In contemplating the foundational architecture of epistemological foundations and propositional truth conditions in formal semantics. one is invariably confronted with a profound historical tension. While early twentieth-century formulations sought strict axiomatic reduction, empirical investigations con...
ML_DOC_002	English	en	Latin	GPT_4o	false	45	2,471	24,692	013d4a1bb88bc60093d11537a7eacd0103a2f21648764c8e8e65dbab0037bdab	Long-Form Scholarly Treatise on Epistemic Systems (English - GPT_4o)	From a computational perspective, analyzing epistemological foundations and propositional truth conditions in formal semantics. requires establishing clear theoretical bounds on algorithmic tractability. Under standard complexity-theoretic assumptions, the optimization surface exhibits structural regularities that perm...
ML_DOC_003	English	en	Latin	Claude_3_5_Sonnet	false	45	3,236	26,312	d5c9369f314b5c574b22762cbfaac3b8ad71a8fb869aa795b3b538807b963240	Long-Form Scholarly Treatise on Epistemic Systems (English - Claude_3_5_Sonnet)	When examining the nuanced implications of epistemological foundations and propositional truth conditions in formal semantics. it is essential to balance mathematical rigor with a reflexive awareness of underlying modeling assumptions. Rather than presuming an absolute isomorphism between formal models and empirical re...
ML_DOC_004	English	en	Latin	Gemini_3_7_Flash	false	45	2,426	22,352	a3f42d813763d3789210308e74e41a2723780337ddb423808e63a16a2946ee51	Long-Form Scholarly Treatise on Epistemic Systems (English - Gemini_3_7_Flash)	The epistemic justification for models of epistemological foundations and propositional truth conditions in formal semantics. hinges upon coherentist verification: theoretical propositions must demonstrate mutual entailment across independent observation channels. Specifically, we delineate three principal criteria: fi...
ML_DOC_005	English	en	Latin	DeepSeek_R1	false	45	2,831	23,432	854bceecfda631df6ce95fa60e36318f0a27a3ab3b1cae0c15593c8c77e8005d	Long-Form Scholarly Treatise on Epistemic Systems (English - DeepSeek_R1)	Considering the Riemannian geometry underlying epistemological foundations and propositional truth conditions in formal semantics.—we observe that the curvature of the loss manifold directly dictates gradient descent dynamics. By computing the spectrum of the Fisher information matrix, one identifies low-dimensional su...
ML_DOC_006	English	en	Latin	Qwen_2_5_72B	false	45	2,786	23,252	f2344757d5ac932f8ca9fe21877529e39e2717ae34bcb2ec2abf1a21ce243fcf	Long-Form Scholarly Treatise on Epistemic Systems (English - Qwen_2_5_72B)	A systematic taxonomy of epistemological foundations and propositional truth conditions in formal semantics. reveals three interrelated methodological tiers. At the macro level, global conservation laws govern aggregate state transitions. At the meso level, modular interactions mediate information exchange between func...
ML_DOC_007	English	en	Latin	Llama_3_3_70B	false	45	2,606	23,252	987c4941afd188a74ee3ae4306a1ff0ded9946a296e47c4b3f3a03be22d69929	Long-Form Scholarly Treatise on Epistemic Systems (English - Llama_3_3_70B)	Investigating epistemological foundations and propositional truth conditions in formal semantics. demonstrates how structural invariants govern macroscopic thermodynamic behavior. By applying statistical mechanics principles to distributed systems, we establish that phase transitions occur when the density of inter-nod...
ML_DOC_008	Spanish	es	Latin	Human_Academic	true	42	2,684	22,271	bd339a2c6d113abe72389b5ca825f0a251c587fccae5e5309a4ba079ce18519d	Long-Form Scholarly Treatise on Epistemic Systems (Spanish - Human_Academic)	Proposición 1. Al examinar la estructura ontológica de epistemological foundations and propositional truth conditions in formal semantics., surge una tensión hermenéutica ineludible en el discurso filosófico contemporáneo. Las reducciones axiomáticas del positivismo lógico clásico resultan manifiestamente insuficientes...
ML_DOC_009	Spanish	es	Latin	GPT_4o	false	45	2,561	21,803	c4ca1bd8f961cfa206b46f5cf479f06cf7d4d8840569541d3921566d1ea69d4f	Long-Form Scholarly Treatise on Epistemic Systems (Spanish - GPT_4o)	"Sección 1. El análisis computacional sistemático de epistemological foundations and propositiona(...TRUNCATED)
ML_DOC_010	Spanish	es	Latin	Claude_3_5_Sonnet	false	45	2,561	21,803	c4ca1bd8f961cfa206b46f5cf479f06cf7d4d8840569541d3921566d1ea69d4f	Long-Form Scholarly Treatise on Epistemic Systems (Spanish - Claude_3_5_Sonnet)	"Sección 1. El análisis computacional sistemático de epistemological foundations and propositiona(...TRUNCATED)

End of preview.

ClozeCongruence 3.0: Multilingual AI Text Forensics & Provenance Benchmark

Official reproducible benchmark datasets accompanying the research paper: "ClozeCongruence 3.0: Cross-Lingual Macro-Discourse Reconstruction and Dynamic Burstiness Gating for Zero-Shot AI Text Forensics across 10 Typologically Diverse Languages" (Debdip Bandyopadhyay, 2026).

Benchmark Splits

This benchmark spans 10 global languages across 6 typological writing scripts:

Latin: English (en), German (de), Spanish (es), French (fr)
Cyrillic: Russian (ru)
CJK: Simplified Chinese (zh), Japanese (ja)
Arabic: Modern Standard Arabic (ar)
Devanagari / Indic: Hindi (hi), Bengali (bn)

Included Datasets

data/multilingual_longform_benchmark.json:
- 70 scholarly documents (42-45 paragraphs each; 3,120 paragraphs, 138,008 words).
- Evaluated across 7 author classes: Human Academic, OpenAI GPT-4o, Anthropic Claude 3.5 Sonnet, Google Gemini 3.7 Flash, DeepSeek-R1, Alibaba Qwen 2.5 72B, and Meta Llama 3.3 70B.
data/real_world_downloaded_human_benchmark.json:
- 10 authentic academic treatises downloaded from Wikipedia across all 10 target languages (447 paragraphs, 19,161 words) establishing real-world 0.00% False Positive Rate (FPR).
data/multilingual_mixed_human_ai_benchmark.json:
- 30 long-form documents across 3 realistic human-AI collaboration regimes (50/50 interleaved, 25/75, and 75/25).

Key Empirical Metrics

Mean 10-Language AUROC: 98.92%
Native Human False Positive Rate (FPR): 0.00% (Zero false accusations)
Top-1 LLM Attribution Accuracy: 99.71%
Cryptographic Provenance: ISO/IEC 27037 Tamper-Evident SHA-256 and Ed25519 digital certificates.

Citation

@article{bandyopadhyay2026cloze3,
  title={ClozeCongruence 3.0: Cross-Lingual Macro-Discourse Reconstruction and Dynamic Burstiness Gating for Zero-Shot AI Text Forensics across 10 Typologically Diverse Languages},
  author={Bandyopadhyay, Debdip},
  journal={Transactions on Machine Learning Research},
  year={2026}
}

Downloads last month: 37

Total file size:

2.73 MB