Differences

This shows you the differences between two versions of the page.

--- nlp:commonsense_reasoning [2021/10/21 21:54] – [Papers] jmflanig
+++ nlp:commonsense_reasoning [2023/07/25 01:07] (current) – [Datasets] jmflanig
@@ Line 2: / Line 2: @@
 ===== Papers =====
+==== General Papers ====
   * [[https://www.aaai.org/Papers/Symposia/Spring/2002/SS-02-09/SS02-09-011.pdf|Singh 2002 - The Public Acquisition of Commonsense Knowledge]] The Open Mind Common Sense project, which was the foundation of ConceptNet
   * [[https://www.aclweb.org/anthology/D14-1059.pdf|Angeli & Manning 2014 - NaturalLI: Natural Logic Inference for Common Sense Reasoning]]
   * [[https://arxiv.org/pdf/1612.03975.pdf|Speer et al 2016 - ConceptNet 5.5: An Open Multilingual Graph of General Knowledge]]
   * [[https://arxiv.org/pdf/1808.05326.pdf|Zellers et al 2018 - SWAG: A Large-Scale Adversarial Dataset for Grounded Commonsense Inference]]
-  * [[https://arxiv.org/pdf/1905.07830.pdf|Zellers et al 2019 - HellaSwag: Can a Machine Really Finish Your Sentence?]]
+    * [[https://arxiv.org/pdf/1905.07830.pdf|Zellers et al 2019 - HellaSwag: Can a Machine Really Finish Your Sentence?]]
-  * [[https://arxiv.org/pdf/1811.00146.pdf|Sag et al 2019 - ATOMIC: An Atlas of Machine Commonsense for If-Then Reasoning]]
+  * [[https://arxiv.org/pdf/1811.00146.pdf|Sag et al 2018 - ATOMIC: An Atlas of Machine Commonsense for If-Then Reasoning]]
+  * [[https://arxiv.org/pdf/1909.00505.pdf|Feldman et al 2019 - Commonsense Knowledge Mining from Pretrained Models]]
+  * [[https://arxiv.org/pdf/1906.05317.pdf|Bosselut et al 2019 - COMET: Commonsense Transformers for Automatic Knowledge Graph Construction]] Can be used as a language model-like interface to knowledge contained in ATOMIC or ConceptNet. The “mask” tokens are padding for batch training, see [[https://github.com/atcbosselut/comet-commonsense/issues/31|here]]. **[[https://mosaickg.apps.allenai.org/model_comet2020_people_events|Demo]]**
+  * [[https://aclanthology.org/2022.acl-long.225.pdf|Liu et al 2022 - Generated Knowledge Prompting for Commonsense Reasoning]]
+==== Commonsense Question Answering ====
+  * [[https://arxiv.org/pdf/1811.00937.pdf|Talmor et al 2018 - CommonsenseQA: A Question Answering Challenge Targeting Commonsense Knowledge]]
   * [[https://arxiv.org/pdf/2004.05483.pdf|Shwartz et al 2020 - Unsupervised Commonsense Question Answering with Self-Talk]]
-  * [[https://arxiv.org/pdf/1906.05317.pdf|Bosselut et al COMET: Commonsense Transformers for Automatic Knowledge Graph Construction]] Can be used as a language model-like interface to knowledge contained in ATOMIC or ConceptNet. The “mask” tokens are padding for batch training, see [[https://github.com/atcbosselut/comet-commonsense/issues/31|here]]. [[https://mosaickg.apps.allenai.org/model_comet2020_people_events|Demo]]
+==== Abductive Reasoning ====
+  * Papers
+    * [[https://arxiv.org/pdf/1908.05739.pdf|Bhagavatula et al 2019 - Abductive Commonsense Reasoning]]
 ===== Applications ======
@@ Line 24: / Line 38: @@
 ===== Datasets =====
+A very helpful list of resources is here: [[https://cs.nyu.edu/~davise/Benchmarks/Text.html]]
   * [[https://rowanzellers.com/swag/|SWAG]]
   * [[https://rowanzellers.com/hellaswag/|HellaSWAG]]
   * Winograd Schema Challenge ([[https://cs.nyu.edu/~davise/papers/WinogradSchemas/WS.html|original dataset]])
+  * [[https://www.tau-nlp.sites.tau.ac.il/commonsenseqa|CommonsenseQA]] [[https://arxiv.org/pdf/1811.00937.pdf|paper]]
+  * Abductive Reasoning Datasets
+    * ART, $\alpha$NLI and $\alpha$NLG: 20k commonsense narrative contexts and 200k explanations (from ROCStories dataset) [[https://arxiv.org/pdf/1908.05739.pdf|paper]]
+===== Demos =====
+  * COMET [[https://mosaickg.apps.allenai.org/model_comet2020_people_events|Demo]]
 ===== Tutorials =====