From 5a22281d61d30fe3f2de6f7b259c03e2c857975d Mon Sep 17 00:00:00 2001
From: Yusheng Su <yushengsu.thu@gmail.com>
Date: Wed, 22 Nov 2023 20:20:20 +0400
Subject: [PATCH] Update about.md

---
 _pages/about.md | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/_pages/about.md b/_pages/about.md
index 2ac63ce34aa29..2bd97c9798468 100644
--- a/_pages/about.md
+++ b/_pages/about.md
@@ -27,7 +27,7 @@ My research spans the areas of natural language processing and machine learning,
 
 <!--Recently, I more focus on <b>interactable and reliable</b> part:-->
 
-* <b>Interactive AI Agent. (LLM Planning and Reasoning)</b> Developing autonomous systems (foundation model agents) that can acquire external information through interaction with the external (environment/human) and utilize the acquired knowledge to drive their decision-making and accomplish more complex long-horizon tasks ([AgentVerse](https://arxiv.org/abs/2308.10848), [XAgent](https://blog.x-agent.net/about/), [CokeBERT](https://arxiv.org/abs/2009.13964), [CSS-LM](https://arxiv.org/abs/2102.03752), [Tool Leaning](https://arxiv.org/abs/2304.08354), [ChatDev](https://arxiv.org/abs/2307.07924)). <small> Note that: We further make autonomous systems (agents) to collaboratively accomplish complex tasks and observe the [emergence of human-like social behaviors](https://arxiv.org/abs/2308.10848). </small>
+* <b>Interactive AI Agent. (LLM Planning and Reasoning)</b> Developing autonomous agents (based on LLMs) that can acquire external information through interaction with the external (environment/human) and utilize the acquired knowledge to drive their decision-making and accomplish more complex long-horizon tasks ([AgentVerse](https://arxiv.org/abs/2308.10848), [XAgent](https://blog.x-agent.net/about/), [CokeBERT](https://arxiv.org/abs/2009.13964), [CSS-LM](https://arxiv.org/abs/2102.03752), [Tool Leaning](https://arxiv.org/abs/2304.08354), [ChatDev](https://arxiv.org/abs/2307.07924)). <small> Note that: I explore how to make multiple autonomous agents collaborate more efficiently to accomplish complex tasks and am further instered in their [emergent human-like social behaviors](https://arxiv.org/abs/2308.10848). </small>
 
 * <b>AI Alignment. (LLM Alignment)</b> Designing paradigms to control, steer, teach, and guide AI systems that smarter than human ([scalable oversight](https://openai.com/blog/introducing-superalignment)). At the present stage, I am studying how to create a tool that can explore unclosed behavior in LLMs ([Model Emotion](https://arxiv.org/abs/2302.09582) and propose a metric/pipeline to autonomously evaluate/supervise LLMs ([Chateval](https://arxiv.org/abs/2308.07201)).