Commit 8d738f52 by Zidong Du

~

parent c26556f4
...@@ -7,6 +7,7 @@ ...@@ -7,6 +7,7 @@
\usepackage{color} \usepackage{color}
\newcommand{\rmk}[1]{\textcolor{red}{--[#1]--}} \newcommand{\rmk}[1]{\textcolor{red}{--[#1]--}}
\newcommand{\note}[1]{\textcolor{red}{#1}} \newcommand{\note}[1]{\textcolor{red}{#1}}
\usepackage{enumitem}
\usepackage{aaai21} % DO NOT CHANGE THIS \usepackage{aaai21} % DO NOT CHANGE THIS
\usepackage{times} % DO NOT CHANGE THIS \usepackage{times} % DO NOT CHANGE THIS
...@@ -166,13 +167,13 @@ ...@@ -166,13 +167,13 @@
\maketitle \maketitle
\begin{abstract} \begin{abstract}
The natural emergence of symbolic languages with high compositionality has The emergence of symbolic languages with high compositionality has
attracted extensive attention from a broad range of communities. Existing attracted extensive attention from a broad range of communities. Existing
studies achieve high compositionality through \emph{deliberately handcrafted} studies achieve high compositionality through \emph{deliberately handcrafted}
inductions (e.g., small vocabulary sizes, carefully constructed distractors, inductions (e.g., small vocabulary sizes, carefully constructed distractors,
and ease-of-teaching) in multi-agent learning, which are unnatural. and ease-of-teaching) in multi-agent learning, which are unnatural.
Yet, few studies investigate the emergence of symbolic language with high Yet, few studies investigate the emergence of symbolic language with high
compositionality \emph{``naturally''}, i.e., without any deliberately handcrafted compositionality \emph{naturally}, i.e., without any deliberately handcrafted
inductions. inductions.
In this paper, we are the first to successfully achieve high compositional symbolic In this paper, we are the first to successfully achieve high compositional symbolic
...@@ -182,11 +183,14 @@ ...@@ -182,11 +183,14 @@
inductions, we observe that the agent capacity plays a key role in inductions, we observe that the agent capacity plays a key role in
compositionality. We further reveal and characterize the quantitative relationship compositionality. We further reveal and characterize the quantitative relationship
between the agent capacity and the compositionality of symbolic language both between the agent capacity and the compositionality of symbolic language both
theoretically and experimentally. The theoretical analysis is built on the MSC theoretically and experimentally.
(Markov Series Channel) model for the language transmission process and a %The theoretical analysis is built on the MSC
novel mutual information-based metric for the compositionality. The %(Markov Series Channel) model for the language transmission process and a
experiments are conducted on a listener-speaker referential game framework %novel mutual information-based metric for the compositionality.
with eliminated external environment factors. Both theoretical analysis and %The
%experiments are conducted on a listener-speaker referential game framework
%with eliminated external environment factors.
Both theoretical analysis and
experimental results lead to a counter-intuitive conclusion that lower agent experimental results lead to a counter-intuitive conclusion that lower agent
capacity facilitates the emergence of symbolic language with higher capacity facilitates the emergence of symbolic language with higher
compositionality. \note{Based on our conclusion, we can generate higher compositionality. \note{Based on our conclusion, we can generate higher
......
Markdown is supported
0% or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment