Updates from Overleaf

chapter-4.tex

@@ -2,3 +2,13 @@ \chapter{Generating Content to Optimize Behavior}
 \label{chatper:Generating Content Leading to Optimal Behavior}
 
 
+In the last chapter, we discussed large content and behavior models (LCBMs) with the ability to generate content conditioned on behavior, generate (simulate) behavior conditioned on content, and understanding of content and behavior. In this chapter, we focus specifically on the capability of generating content conditioned on behavior. LCBMs were built following the instruction fine-tuning paradigm. Using LCBMs, we showed that including behavior data as receiver tokens along with content data (communicator tokens) helps complete the entire communication flow and train the LLM to teach it both the receiver side and the communicator side of the flow. 
+
+
+In this chapter, we take a deeper look into the common use case of generating content which can help get the behavior the communicator wants. For instance, a marketer wants to write emails or compose tweets that will bring her the maximum number of link clicks and likes. We propose several solutions to solve this problem and compare several paradigms to achieve this. We show this over both short-term key performance indicators (downloads and likes), and long-term indicators (brand and content memorability). 
+
+
+
+\section{}
+
+

main.tex

@@ -203,10 +203,11 @@ \chapter{Introduction: The Two Cultures of Behavioral Sciences}
 
 \end{comment}
 
+A \textit{modality} is defined in terms of information, such that a modality is a medium through which information is conveyed \cite{liang2022foundations,grifoni2009multimodal,martin2001annotation}. Similarly, a multimodal distribution is defined as having more than one peak in the probability distribution describing the nature of information. Behavior as a modality occurs in the process of communication. Communication includes all of the procedures by which one mind may affect another \cite{shannon-weaver-1949}. This includes all forms of expression, such as words, gestures, speech, pictures, and musical sounds. Communication can be seen as being composed of seven modalities (Fig.~\ref{fig:factors-of-communication}): (the communicator, message, time of message, channel, receiver, time of effect, and effect). These modalities can vary independently of each other \cite{khandelwal2023large,khurana2023behavior,si2023long} but carry signals about each other \cite{khurana-etal-2023-synthesizing}. The message as a modality carries data from the communicator to receiver and encodes information generated by the communicator. Similarly, behavior as a modality carries data from the receiver and encodes information generated by the receiver. This is often a continuous cycle, where behavior generated in the previous cycle becomes the message of the next cycle. 
 
 
-Communication includes all of the procedures by which one mind may affect another \cite{shannon-weaver-1949}. This includes all forms of expression, such as words, gestures, speech, pictures, and musical sounds.
-Communication can be seen as being composed of seven parts (Fig.~\ref{fig:factors-of-communication}): (the communicator, message, time of message, channel, receiver, time of effect, and effect). Different fields deal with different parts of behavior . I will give a broad overview of these fields in the upcoming paragraphs, but two streams have emerged broadly in behavioral sciences: explanation and prediction of behavior (receiver effect) \cite{breiman2001statistical,hofman2017prediction,shmueli2010explain}. 
+
+Different fields deal with different parts of behavior. I will give a broad overview of these fields in the upcoming paragraphs, but two streams have emerged broadly in behavioral sciences: explanation and prediction of behavior (receiver effect) \cite{breiman2001statistical,hofman2017prediction,shmueli2010explain}. 
 
 
 \begin{figure*}[!t]
@@ -1518,11 +1519,11 @@ \chapter*{Publications}
 
  \item Kumar, Y., Jha, R., Gupta, A., Aggarwal, M., Garg, A., Malyan, T., Bhardwaj, A., Ratn Shah, R., Krishnamurthy, B., \& Chen, C. (2023). Persuasion Strategies in Advertisements. Proceedings of the AAAI Conference on Artificial Intelligence, 37(1), 57-66. \url{https://doi.org/10.1609/aaai.v37i1.25076}
 
- \item Bhattacharya, A., Singla, Y. K., Krishnamurthy, B., Shah, R. R., \& Chen, C. (2023). A Video Is Worth 4096 Tokens: Verbalize Videos To Understand Them In Zero Shot. In Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing, pages 9822–9839, Singapore. Association for Computational Linguistics. (Nominated for the best paper award!)
+ \item Bhattacharya, A., Singla, Y. K., Krishnamurthy, B., Shah, R. R., \& Chen, C. (2023). A Video Is Worth 4096 Tokens: Verbalize Videos To Understand Them In Zero Shot. In Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing, pages 9822–9839, Singapore. Association for Computational Linguistics. (\textbf{Nominated for the best paper award!})
 
- \item Khandelwal, A., Agrawal, A., Bhattacharyya, A., Singla, Y.K., Singh, S., Bhattacharya, U., Dasgupta, I., Petrangeli, S., Shah, R.R., Chen, C. and Krishnamurthy, B., 2024. Large Content And Behavior Models To Understand, Simulate, And Optimize Content And Behavior. International Conference on Learning Representations. 
+ \item Khandelwal, A., Agrawal, A., Bhattacharyya, A., Singla, Y.K., Singh, S., Bhattacharya, U., Dasgupta, I., Petrangeli, S., Shah, R.R., Chen, C. and Krishnamurthy, B., 2024. Large Content And Behavior Models To Understand, Simulate, And Optimize Content And Behavior. International Conference on Learning Representations. (\textbf{Spotlight and nominated for award!})
 
- \item S I, H., Singh, S., K Singla, Y., Krishnamurthy, B., Chen, C., Baths V., \& Ratn Shah, R. (2023). Sharingan: How Much Will Your Customers Remember Your Brands After Seeing Your Ads?. arxiv preprint (Under review).
+ \item S I, H., Singh, S., K Singla, Y., Krishnamurthy, B., Chen, C., Baths V., \& Ratn Shah, R. (2024). Long-Term Ad Memorability: Understanding and Generating Memorable Ads. arxiv preprint (Under review).
 
  \item Khurana, V., Singla, Y.K., Subramanian, J., Shah, R.R., Chen, C., Xu, Z. and Krishnamurthy, B., 2023. Behavior Optimized Image Generation. arXiv preprint arXiv:2311.10995. (Under review)
 

ref.bib

@@ -3855,6 +3855,22 @@ @inproceedings{joshi-etal-2015-harnessing
  url = "https://aclanthology.org/P15-2124",
  doi = "10.3115/v1/P15-2124",
  pages = "757--762",
+}@inproceedings{martin2001annotation,
+ title={On the annotation of the multimodal behavior and computation of cooperation between modalities},
+ author={Martin, Jean-Claude and Grimard, Sarah and Alexandri, Katerina},
+ booktitle={Proceedings of the Workshop on Multimodal Communicatiom and Context in Embodied Agents, Fifth International Conference on Autonomous Agents},
+ pages={1--7},
+ year={2001}
+}@book{grifoni2009multimodal,
+ title={Multimodal human computer interaction and pervasive services},
+ author={Grifoni, Patrizia},
+ year={2009},
+ publisher={IGI Global}
+}@article{liang2022foundations,
+ title={Foundations and recent trends in multimodal machine learning: Principles, challenges, and open questions},
+ author={Liang, Paul Pu and Zadeh, Amir and Morency, Louis-Philippe},
+ journal={arXiv preprint arXiv:2209.03430},
+ year={2022}
 }
 @article{bai2022training,
  title={Training a Helpful and Harmless Assistant with Reinforcement Learning from Human Feedback},