paper update

Author: thestephencasper

_includes/team.html

@@ -5,7 +5,7 @@ <h3>Principal Investigator</h3>
 
 <p>
     <img style="padding-right: 15px;" src="/docs/assets/dylan.png" width="160" height="160" alt="Dylan Hadfield-Menell">
-    <strong>Dylan Hadfield-Menell</strong>>, <a href="dhm@csail.mit.edu">dhm@csail.mit.edu</a>, <a href="http://people.csail.mit.edu/dhm/">Website</a>
+    <strong>Dylan Hadfield-Menell</strong>, <a href="dhm@csail.mit.edu">dhm@csail.mit.edu</a>, <a href="http://people.csail.mit.edu/dhm/">Website</a>
     <br>
     Dylan is an assistant professor on the faculty of Artificial Intelligence and Decision-Making in the EECS Department and Computer Science and Artificial Intelligence Laboratory (CSAIL) at the Massachusetts Institute of Technology (MIT). His research focuses on the problem of agent alignment: the challenge of identifying behaviors that are consistent with the goals of another actor or group of actors. His work aims to identify algorithmic solutions to alignment problems that arise from groups of AI systems, principal-agent pairs (i.e., human-robot teams), and societal oversight of ML systems.
 </p>

index.markdown

@@ -7,7 +7,7 @@ permalink: /
 
 ## Home
 
-Our focus is to better align the development of AI with human interests and values headed by [Dylan Hadfield-Menell](https://engineering.mit.edu/faculty/dylan-hadfield-menell/). We are working in the [Embodied Entelligence Group](https://ei.csail.mit.edu/) at [MIT CSAIL](https://www.csail.mit.edu/) to build the conceptual understanding and algorithmic techniques that will be needed for more trustworthy AI. Our research is interdisciplinary but emphasizes how humans and AI systems interact in the contexts of value learning, incentives, recommendation, and debugging.
+Our focus is to better align the development of AI with human interests and values headed by [Dylan Hadfield-Menell](http://people.csail.mit.edu/dhm/). We are working in the [Embodied Entelligence Group](https://ei.csail.mit.edu/) at [MIT CSAIL](https://www.csail.mit.edu/) to build the conceptual understanding and algorithmic techniques that will be needed for more trustworthy AI. Our research is interdisciplinary but emphasizes how humans and AI systems interact in the contexts of value learning, incentives, recommendation, and debugging.
 
 
 

research.markdown

@@ -7,10 +7,10 @@ permalink: /research/
 
 ## Research
 
-Find us on [Github](https://github.com/Algorithmic-Alignment-Lab).   
+Find us on [Github](https://github.com/Algorithmic-Alignment-Lab).
 
-<!--- #### 2022
+###2022
 
-Sutton, R. S., & Barto, A. G. (2018). [Reinforcement learning: An introduction.](https://web.stanford.edu/class/psych209/Readings/SuttonBartoIPRLBook2ndEd.pdf) MIT press. [BibTeX](https://scholar.googleusercontent.com/scholar.bib?q=info:t8N5xiW9bXoJ:scholar.google.com/&output=citation&scisdr=CgWTY31kEPyMgqtR1ow:AAGBfm0AAAAAYiBXzowXsFVJNPzBvJ5wFyFoi8IN8GkG&scisig=AAGBfm0AAAAAYiBXztuv9gUZtgxBLqD3ECitmd9rQZAc&scisf=4&ct=citation&cd=-1&hl=en) --->
+Casper, S., Nadeau, M., Hadfield-Menell, D, & Kreiman, G (2022). [Robust Feature-Level Adversaries are Interpretability Tools](https://arxiv.org/abs/2110.03605). [BibTeX](https://dblp.uni-trier.de/rec/journals/corr/abs-2110-03605.html?view=bibtex)