mlsanityData to 14 Oct 2026

PInKS: Preconditioned Commonsense Inference with Minimal Supervision

arXiv 2206.07920 ↗First posted 16 Jun 2022cs.CLcited by 4

Not on the map. Why?

Abstract (arXiv)

Reasoning with preconditions such as "glass can be used for drinking water unless the glass is shattered" remains an open problem for language models. The main challenge lies in the scarcity of preconditions data and the model's lack of support for such reasoning. We present PInKS, Preconditioned Commonsense Inference with WeaK Supervision, an improved model for reasoning with preconditions through minimum supervision. We show, both empirically and theoretically, that PInKS improves the results on benchmarks focused on reasoning with the preconditions of commonsense knowledge (up to 40% Macro-F1 scores). We further investigate PInKS through PAC-Bayesian informativeness analysis, precision measures, and ablation study.

Key facts. Cited by 4 papers in the map (data through 14 Oct 2026). Builds on 39 of its 73 references that are in the map. First posted on arXiv 16 Jun 2022.

Builds on

39 of 73 references.

References · 39

  1. 2019RoBERTa: A Robustly Optimized BERT Pretraining Approachcited by 13,372
  2. 2017A Broad-Coverage Challenge Corpus for Sentence Understanding through Inferencecited by 2,682
  3. 2019Transformer-XL: Attentive Language Models Beyond a Fixed-Length Contextcited by 1,850
  4. 2015A large annotated corpus for learning natural language inferencecited by 1,847
  5. 2018CommonsenseQA: A Question Answering Challenge Targeting Commonsense Knowledgecited by 1,822
  6. 2019SuperGLUE: A Stickier Benchmark for General-Purpose Language Understanding Systemscited by 1,729
  7. 2019VisualBERT: A Simple and Performant Baseline for Vision and Languagecited by 1,296
  8. 2016ConceptNet 5.5: An Open Multilingual Graph of General Knowledgecited by 1,067
  9. 2017Virtual Adversarial Training: A Regularization Method for Supervised and Semi-Supervised Learningcited by 920
  10. 2016A Corpus and Evaluation Framework for Deeper Understanding of Commonsense Storiescited by 417
  11. 2018Tune: A Research Platform for Distributed Model Selection and Trainingcited by 395
  12. 2017Snorkel: Rapid Training Data Creation with Weak Supervisioncited by 372
27 more
  1. 2020COMET-ATOMIC 2020: On Symbolic and Neural Commonsense Knowledge Graphscited by 254
  2. 2016Data Programming: Creating Large Training Sets, Quicklycited by 246
  3. 2019KagNet: Knowledge-Aware Graph Networks for Commonsense Reasoningcited by 216
  4. 2018Combining Fact Extraction and Verification with Neural Semantic Matching Networkscited by 120
  5. 2019ASER: A Large-scale Eventuality Knowledge Graphcited by 89
  6. 2020GLUCOSE: GeneraLized and COntextualized Story Explanationscited by 86
  7. 2017HoloClean: Holistic Data Repairs with Probabilistic Inferencecited by 79
  8. 2016Label-Free Supervision of Neural Networks with Physics and Domain Knowledgecited by 74
  9. 2014Curriculum Learning of Multiple Taskscited by 66
  10. 2017Neural Ranking Models with Weak Supervisioncited by 61
  11. 2020Transfer Learning and Distant Supervision for Multilingual Transformer Models: A Study on African Languagescited by 52
  12. 2017Learning the Structure of Generative Models without Labeled Datacited by 50
  13. 2021Distantly-Supervised Named Entity Recognition with Noise-Robust Learning and Language Model Augmented Self-Trainingcited by 41
  14. 2020Temporal Common Sense Acquisition with Minimal Supervisioncited by 39
  15. 2021Ultra-Fine Entity Typing with Weak Supervision from a Masked Language Modelcited by 39
  16. 2020Facts as Experts: Adaptable and Interpretable Neural Memory over Symbolic Knowledgecited by 37
  17. 2021Lawyers are Dishonest? Quantifying Representational Harms in Commonsense Knowledge Resourcescited by 30
  18. 2020Learning to Rationalize for Nonmonotonic Reasoning with Distant Supervisioncited by 27
  19. 2021"I'm Not Mad": Commonsense Implications of Negation and Contradictioncited by 26
  20. 2019Question Answering is a Format; When is it Useful?cited by 25
  21. 2020Advanced Semantics for Commonsense Knowledge Extractioncited by 24
  22. 2021skweak: Weak Supervision Made Easy for NLPcited by 17
  23. 2021PaCo: Preconditions Attributed to Commonsense Knowledgecited by 9
  24. 2021Table-based Fact Verification with Salience-aware Learningcited by 9
  25. 2020Modeling Preconditions in Text with a Crowd-sourced Datasetcited by 8
  26. 2020Foreseeing the Benefits of Incidental Supervisioncited by 7
  27. 2021Randomized Deep Structured Prediction for Discourse-Level Processingcited by 4

Built on by

Cited by year

Citing papers per year: 2023: 2; 2024: 2.

4 papers.

  1. 2024Making Large Language Models into World Models with Precondition and Effect Knowledgecited by 22
  2. 2023Traffic-Domain Video Question Answering with Automatic Captioningcited by 10
  3. 2024TEXT2AFFORD: Probing Object Affordance Prediction abilities of Language Models solely from Textcited by 3
  4. 2023Preconditioned Visual Language Inference with Weak Supervisioncited by 0