datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
proofwriter
Dataset Card for "proofwriter"
More Information needed
proofwriter_processed_OWAproofwriter
Dataset Card for "proofwriter"
More Information needed
proofwriter-source
ProofWriter (The Source)
An unmodified copy of AI2's ProofWriter dataset (release V2020.12.3), re-hosted as datasets
configs for convenient loading. The records are faithful to the upstream release — the id-keyed JSON
is preserved as-is; typing and reasoning-graph extraction happen in later stages.
Each config is a {world}-depth-{n} shelf of the synthetic core (OWA/CWA × depths
0/1/2/3/5), split train/dev/test (dev kept as the corpus names it).
Source:… See the full description on the dataset page: https://huggingface.co/datasets/arqa39/proofwriter-source.proofwriter-deduction-balancedA processed subset of the OWA section of the ProofWriter dataset.
Each train/test split contains 300 entries, each of which has a unique set of theories and a single question for those theories.
Both splits are balanced so that the depth of the proof required to answer the question varies evenly between 0-5 (50 entries each), and the labels are balanced (100 each).
'Unknown' labels have been replaced by 'Uncertain' to match other datasets.
proofwriter-source
ProofWriter (The Source)
An unmodified copy of AI2's ProofWriter dataset (release V2020.12.3), re-hosted as datasets
configs for convenient loading. The records are faithful to the upstream release — the id-keyed JSON
is preserved as-is; typing and reasoning-graph extraction happen in later stages.
Each config is a {world}-depth-{n} shelf of the synthetic core (OWA/CWA × depths
0/1/2/3/5), split train/dev/test (dev kept as the corpus names it).
Source:… See the full description on the dataset page: https://huggingface.co/datasets/rlhf-and-friends/proofwriter-source.ProofWriter
Github
https://github.com/teacherpeterpan/Logic-LLM/blob/main/outputs/logic_programs/ProofWriter_dev_gpt-4.json
Reference
@inproceedings{PanLogicLM23,
author = {Liangming Pan and
Alon Albalak and
Xinyi Wang and
William Yang Wang},
title = {{Logic-LM:} Empowering Large Language Models with Symbolic Solvers for Faithful Logical Reasoning},
booktitle = {Findings of the 2023 Conference on Empirical… See the full description on the dataset page: https://huggingface.co/datasets/renma/ProofWriter.proofwriter
ProofWriter — structured
A cleaned, structured build of AI2's ProofWriter
for logical entailment with reasoning-graph supervision. Each row is one theory
(facts + Horn-clause rules) with the questions posed against it; the zip's formal string
reps are parsed into typed atoms (subject, relation, object, polarity), and every
question keeps its gold answer and gold derivation as a structured proof graph — so no
natural-language reverse-engineering is needed downstream.… See the full description on the dataset page: https://huggingface.co/datasets/rlhf-and-friends/proofwriter.proofwriter-datasetproofwriter-mirror
ProofWriter (The Mirror)
A typed, content-faithful mirror of AI2's ProofWriter dataset (release V2020.12.3),
derived from proofwriter-source. The JSON encoding is cleaned up: the id-keyed dicts
(triple1, Q3, …) become lists of structs that keep their id, every atom representation
is parsed into a typed {subject, relation, object, polarity} triple, and the closed enums
(answer, strategy) are typed. The content stays faithful — nothing renamed, no rows
dropped — and the recursive… See the full description on the dataset page: https://huggingface.co/datasets/rlhf-and-friends/proofwriter-mirror.proofwriter-mirror
ProofWriter (The Mirror)
A typed, content-faithful mirror of AI2's ProofWriter dataset (release V2020.12.3),
derived from proofwriter-source. The JSON encoding is cleaned up: the id-keyed dicts
(triple1, Q3, …) become lists of structs that keep their id, every atom representation
is parsed into a typed {subject, relation, object, polarity} triple, and the closed enums
(answer, strategy) are typed. The content stays faithful — nothing renamed, no rows
dropped — and the recursive… See the full description on the dataset page: https://huggingface.co/datasets/alexdeath53/proofwriter-mirror.proofwriter-source
ProofWriter (The Source)
An unmodified copy of AI2's ProofWriter dataset (release V2020.12.3), re-hosted as datasets
configs for convenient loading. The records are faithful to the upstream release — the id-keyed JSON
is preserved as-is; typing and reasoning-graph extraction happen in later stages.
Each config is a {world}-depth-{n} shelf of the synthetic core (OWA/CWA × depths
0/1/2/3/5), split train/dev/test (dev kept as the corpus names it).
Source:… See the full description on the dataset page: https://huggingface.co/datasets/alexdeath53/proofwriter-source.proofwriter-mirror
ProofWriter (The Mirror)
A typed, content-faithful mirror of AI2's ProofWriter dataset (release V2020.12.3),
derived from proofwriter-source. The JSON encoding is cleaned up: the id-keyed dicts
(triple1, Q3, …) become lists of structs that keep their id, every atom representation
is parsed into a typed {subject, relation, object, polarity} triple, and the closed enums
(answer, strategy) are typed. The content stays faithful — nothing renamed, no rows
dropped — and the recursive… See the full description on the dataset page: https://huggingface.co/datasets/arqa39/proofwriter-mirror.proofwriter
Dataset Card for Dataset Name
Standard proofwriter dataset as grabbed from LogicLM github. Chain of though has been added.
Dataset Details
Dataset Description
Curated by: [More Information Needed]
Funded by [optional]: [More Information Needed]
Shared by [optional]: [More Information Needed]
Language(s) (NLP): [More Information Needed]
License: [More Information Needed]
Dataset Sources [optional]
Repository: [More Information Needed]
Paper… See the full description on the dataset page: https://huggingface.co/datasets/smoorsmith/proofwriter.ProofWriterMechanisticProbe_ProofWriter_ARC
MechanisticProbe
Processed data for Towards a Mechanistic Interpretation of Multi-Step Reasoning Capabilities of Language Modelsproofwriter_LINC
Dataset Card for Dataset Name
Standard proofwriter dataset in the form necessary to run the LINC code.
Dataset Details
Dataset Description
Curated by: [More Information Needed]
Funded by [optional]: [More Information Needed]
Shared by [optional]: [More Information Needed]
Language(s) (NLP): [More Information Needed]
License: [More Information Needed]
Dataset Sources [optional]
Repository: [More Information Needed]
Paper [optional]: [More… See the full description on the dataset page: https://huggingface.co/datasets/smoorsmith/proofwriter_LINC.proofwriter_gold_formalizedProofWriterproofwriter-premise-grounding-hard-negatives-v1proofwriter-qwen25-7b-rar-delta
proofwriter-qwen25-7b-rar-delta
Per-question delta_RaR annotations on ProofWriter:
how much a model's own rephrasing of a logic problem improves its ability to solve it.
Computed with Qwen/Qwen2.5-7B-Instruct, 30,000 queries at k=32 samples per side.
The "teacher" is not a stronger model and does not think longer. It is the same frozen model
answering the same question, with one extra thing in context — a rephrasing it generated itself.
How each row is produced… See the full description on the dataset page: https://huggingface.co/datasets/nick-rui/proofwriter-qwen25-7b-rar-delta.proofwriter-datasetproofwriter___2txt___Qwen2.5_Math_7B_Instruct___Qwen2.5_Coder_7B_Instructproofwriter___2txt___Qwen2.5_7B_Instruct___Qwen2.5_Coder_7B_Instructproofwriter___3txt___Qwen2.5_7B_Instruct___Qwen2.5_Math_7B_Instruct___Qwen2.5_Coder_7B_Instructstream-2-proofwriterproofwriter___txt___Qwen3-8B
Dataset Card for Dataset Name
Contains text reasoning chain from Qwen3-8B from standard prompt:
[
{
"role": "system",
"content": "Your input fields are:\n1. `context` (str): facts here are assumed to be true\n2. `question` (str)\nYour output fields are:\n1. `reasoning` (str)\n2. `answer` (str): must be one of: True, False, Unknown\nAll interactions will be structured in the following way, with the appropriate values filled in.\n\n[[ ## context ## ]]\n{context}\n\n[[ ##… See the full description on the dataset page: https://huggingface.co/datasets/smoorsmith/proofwriter___txt___Qwen3-8B.proofwriter___2txt___Qwen2.5_7B_Instruct___Qwen2.5_Math_7B_Instructproofwriter-deduction-balancedA processed subset of the OWA section of the ProofWriter dataset.
Each train/test split contains 300 entries, each of which has a unique set of theories and a single question for those theories.
Both splits are balanced so that the depth of the proof required to answer the question varies evenly between 0-5 (50 entries each), and the labels are balanced (100 each).
'Unknown' labels have been replaced by 'Uncertain' to match other datasets.
ProofWriter_formatted_v2