introduce lecture 7

jphall663 · Jun 24, 2023 · f6fb87d · f6fb87d
1 parent dce6b28
commit f6fb87d
Show file tree

Hide file tree

Showing 13 changed files with 760 additions and 0 deletions.
diff --git a/img/GPT_Adv_Prmpt3_crop.jpg b/img/GPT_Adv_Prmpt3_crop.jpg
diff --git a/img/NIST_RMF_img1.png b/img/NIST_RMF_img1.png
diff --git a/img/Superv_ML.png b/img/Superv_ML.png
diff --git a/img/apply_benchmark.png b/img/apply_benchmark.png
diff --git a/img/buzzer.png b/img/buzzer.png
diff --git a/img/defcon.jpg b/img/defcon.jpg
diff --git a/img/engage.png b/img/engage.png
diff --git a/img/pastincidents.jpg b/img/pastincidents.jpg
diff --git a/img/security.png b/img/security.png
diff --git a/img/uncertainty.jpg b/img/uncertainty.jpg
diff --git a/tex/lecture_7.bib b/tex/lecture_7.bib
@@ -0,0 +1,69 @@
+@article{hasan2022algorithmic,
+  title={Algorithmic {B}ias and {R}isk {A}ssessments: {L}essons from {P}ractice},
+  author={Hasan, Ali and Brown, Shea and Davidovic, Jovana and Lange, Benjamin and Regan, Mitt},
+  journal={Digital Society},
+  volume={1},
+  number={2},
+  pages={14},
+  year={2022},
+  publisher={Springer},
+  note={URL: \url{https://philpapers.org/archive/HASABA.pdf}}
+}
+
+@article{atherton2023language,
+  title={The {L}anguage of {T}rustworthy {AI}: {A}n {I}n-{D}epth {G}lossary of {T}erms},
+  author={Atherton, Daniel and Schwartz, Reva and Fontana, Peter and Hall, Patrick},
+  year={2023},
+  publisher={National Institute of Standards and Technology, Gaithersburg, MD},
+  note={URL: \url{https://airc.nist.gov/AI_RMF_Knowledge_Base/Glossary.}}
+}
+
+@misc{iqtlabs,
+  title={A{I} {A}ssurance {A}udit of {R}o{BERT}a, an {O}pen source, {P}retrained {L}arge {L}anguage {M}odel},
+  author={Brennen, Andrea and Ashley, Ryan and Calix, Ricardo and Ben-Joseph, JJ and Sieniawski, George and Gogia, Mona and BNH.AI},
+  year={2022},
+  publisher={IQT Labs},
+  url={https://assets.iqt.org/pdfs/IQTLabs_RoBERTaAudit_Dec2022_final.pdf/web/viewer.html}
+}
+
+@misc{Adversa,
+  title={{T}rusted {AI} {B}log (Series)},
+  author={Adversa.ai},
+  year={2022-2023},
+  publisher={"ADMIN"},
+  url={https://adversa.ai/topic/trusted-ai-blog/}
+}
+
+@misc{prompt_injection,
+  doi = {10.48550/ARXIV.2302.12173},
+  url = {https://arxiv.org/abs/2302.12173},
+  author = {Greshake, Kai and Abdelnabi, Sahar and Mishra, Shailesh and Endres, Christoph and Holz, Thorsten and Fritz, Mario},
+  keywords = {Cryptography and Security (cs.CR), Artificial Intelligence (cs.AI), Computation and Language (cs.CL), Computers and Society (cs.CY), FOS: Computer and information sciences, FOS: Computer and information sciences},
+  title = {More than you've asked for: A Comprehensive Analysis of Novel Prompt Injection Threats to Application-Integrated Large Language Models},
+  publisher = {arXiv},
+  year = {2023}
+}
+
+@inproceedings{weidinger2022taxonomy,
+  title={Taxonomy of {R}isks {P}osed by {L}anguage {M}odels},
+  author={Weidinger, Laura and Uesato, Jonathan and Rauh, Maribeth and Griffin, Conor and Huang, Po-Sen and Mellor, John and Glaese, Amelia and Cheng, Myra and Balle, Borja and Kasirzadeh, Atoosa and others},
+  booktitle={2022 ACM Conference on Fairness, Accountability, and Transparency},
+  pages={214--229},
+  year={2022}
+}
+
+@article{mishra2020dqi,
+  title={{DQI}: Measuring data quality in {NLP}},
+  author={Mishra, Swaroop and Arunkumar, Anjana and Sachdeva, Bhavdeep and Bryan, Chris and Baral, Chitta},
+  journal={arXiv preprint arXiv:2005.00816},
+  year={2020}
+}
+
+@article{schwartz2022towards,
+  title={Towards a {S}tandard for {I}dentifying and {M}anaging {B}ias in {A}rtificial {I}ntelligence},
+  author={Schwartz, Reva and Vassilev, Apostol and Greene, Kristen and Perine, Lori and Burt, Andrew and Hall, Patrick and others},
+  journal={NIST Special Publication},
+  volume={1270},
+  pages={1--77},
+  year={2022}
+}
diff --git a/tex/lecture_7.pdf b/tex/lecture_7.pdf