diff --git a/1907.02893v3.pdf b/1907.02893v3.pdf
new file mode 100644
index 0000000000000000000000000000000000000000..12780fc515814b73bc00f915cf16b2b335bcfd89
Binary files /dev/null and b/1907.02893v3.pdf differ
diff --git a/2301.06627v3.pdf b/2301.06627v3.pdf
new file mode 100644
index 0000000000000000000000000000000000000000..e2150fce1f6d868b7312e761ccf01149988a9968
Binary files /dev/null and b/2301.06627v3.pdf differ
diff --git a/BF00114844.pdf b/BF00114844.pdf
new file mode 100644
index 0000000000000000000000000000000000000000..c8df7735f4e2ccfc32e999961ba6832df8052921
Binary files /dev/null and b/BF00114844.pdf differ
diff --git a/F&P1.pdf b/F&P1.pdf
new file mode 100644
index 0000000000000000000000000000000000000000..bcaa3725addda1d80510f3a3b6e495bb02566427
Binary files /dev/null and b/F&P1.pdf differ
diff --git a/PDP.pdf b/PDP.pdf
new file mode 100644
index 0000000000000000000000000000000000000000..eadac4985915bc42dd6fc7d37342bf0a956b3076
Binary files /dev/null and b/PDP.pdf differ
diff --git a/amu-literature.bib b/amu-literature.bib
index 8ce1b24ee9803bd684374bc9b0f846f4cfb32532..8a6fcf7d826cca370f895e0a600002cefb65f607 100644
--- a/amu-literature.bib
+++ b/amu-literature.bib
@@ -39,7 +39,7 @@
   abstract  = {Natural language is characterized by compositionality: the meaning of a complex expression is constructed from the meanings of its constituent parts. To facilitate the evaluation of the compositional abilities of language processing architectures, we introduce COGS, a semantic parsing dataset based on a fragment of English. The evaluation portion of COGS contains multiple systematic gaps that can only be addressed by compositional generalization; these include new combinations of familiar syntactic structures, or new combinations of familiar words and familiar structures. In experiments with Transformers and LSTMs, we found that in-distribution accuracy on the COGS test set was near-perfect (96{--}99{\%}), but generalization accuracy was substantially lower (16{--}35{\%}) and showed high sensitivity to random seed (+-6{--}8{\%}). These findings indicate that contemporary standard NLP models are limited in their compositional generalization capacity, and position COGS as a good way to measure progress.},
   doi       = {10.18653/v1/2020.emnlp-main.731},
   file      = {:COGS-2020.pdf:PDF},
-  groups    = {data},
+  groups    = {data, Groupe de lecture},
   url       = {https://aclanthology.org/2020.emnlp-main.731},
 }
 
@@ -1110,6 +1110,63 @@
   url     = {https://api.semanticscholar.org/CorpusID:265157937},
 }
 
+@Article{rumelhart1986general,
+  author    = {Rumelhart, David E and Hinton, Geoffrey E and McClelland, James L and others},
+  journal   = {Parallel distributed processing: Explorations in the microstructure of cognition},
+  title     = {A general framework for parallel distributed processing},
+  year      = {1986},
+  number    = {45-76},
+  pages     = {26},
+  volume    = {1},
+  file      = {:PDP.pdf:PDF},
+  groups    = {Groupe de lecture},
+  publisher = {Cambridge, MA: MIT Press},
+}
+
+@Article{elman1991distributed,
+  author    = {Elman, Jeffrey L},
+  journal   = {Machine learning},
+  title     = {Distributed representations, simple recurrent networks, and grammatical structure},
+  year      = {1991},
+  pages     = {195--225},
+  volume    = {7},
+  file      = {:BF00114844.pdf:PDF},
+  groups    = {Groupe de lecture},
+  publisher = {Springer},
+}
+
+@Article{mahowald2024dissociating,
+  author    = {Mahowald, Kyle and Ivanova, Anna A and Blank, Idan A and Kanwisher, Nancy and Tenenbaum, Joshua B and Fedorenko, Evelina},
+  journal   = {Trends in Cognitive Sciences},
+  title     = {Dissociating language and thought in large language models},
+  year      = {2024},
+  file      = {:2301.06627v3.pdf:PDF},
+  groups    = {Groupe de lecture},
+  publisher = {Elsevier},
+}
+
+@Article{fodor1988connectionism,
+  author    = {Fodor, Jerry A and Pylyshyn, Zenon W},
+  journal   = {Cognition},
+  title     = {Connectionism and cognitive architecture: A critical analysis},
+  year      = {1988},
+  number    = {1-2},
+  pages     = {3--71},
+  volume    = {28},
+  file      = {:F&P1.pdf:PDF},
+  groups    = {Groupe de lecture},
+  publisher = {Elsevier},
+}
+
+@Article{arjovsky2019invariant,
+  author  = {Arjovsky, Martin and Bottou, L{\'e}on and Gulrajani, Ishaan and Lopez-Paz, David},
+  journal = {arXiv preprint arXiv:1907.02893},
+  title   = {Invariant risk minimization},
+  year    = {2019},
+  file    = {:1907.02893v3.pdf:PDF},
+  groups  = {Groupe de lecture},
+}
+
 @Comment{jabref-meta: databaseType:bibtex;}
 
 @Comment{jabref-meta: grouping:
diff --git a/jurafsky_martin.pdf b/jurafsky_martin.pdf
index 4b9985694ca0726ad7f1ca590e31dd70ac12aa68..177654198d88db1170616bcfbb075d212904d03f 100644
Binary files a/jurafsky_martin.pdf and b/jurafsky_martin.pdf differ