diff --git a/app/teaching/page.tsx b/app/teaching/page.tsx
index b77463b..055d95f 100644
--- a/app/teaching/page.tsx
+++ b/app/teaching/page.tsx
@@ -44,7 +44,7 @@ const courses: Course[] = [
   {
     id: "CPSC 483/583",
     intro:
-      "Graph structure emerges in many important domain applications, including but not limited to natural sciences, social networks, language, vision and database. This course offers a guide to model real-world data in the form of graphs and apply deep learning algorithms to perform predictive and generative AI tasks. The first part of the course is an introduction to representation learning for graphs, and covers thoery and algorithms in the field, including distributed node embeddings, graph neural networks, deep graph generative models, knowledge graphs and non-Euclidean embeddings. The second part touches upon topics of recent interests including graph Transformers and graph learning explainability. The second part of the course also focuses on the emerging field of graph foundation models. Additionally, the course covers important applications of graph machine learning. We learn ways to model data as graphs and apply graph learning techniques to problems in domains including online recommender systems, knowledge graphs, biological networks, molecular structure and physical simulations. The course covers many deep techniques (graph neural networks, graph deep generative models) catered to graph structures. The course assumes background in basic deep learning and PyTorch programming. We will provide a brief basic deep learning tutorial in this course.",
+      "This course provides an introduction to machine learning algorithms designed for graph-structured data, such as social networks, knowledge graphs, and molecular structures. The course covers key graph representation learning techniques, such as node embeddings, graph neural networks, deep generative models for graphs, and non-Euclidean embeddings and other practical topics like AutoML and explainability in graph learning. We also explore applications of graph machine learning, showing how to model and apply graph learning techniques to areas such as online recommender systems, knowledge graphs, biological networks, and physical simulations.",
     happens: [
       {
         year: "2024 Fall",
@@ -60,4 +60,15 @@ const courses: Course[] = [
       },
     ],
   },
+  {
+    id: "CPSC 471/571",
+    intro:
+      "This course provides an in-depth exploration of the principles and practices required to build reliable, fair, and secure machine learning systems. As machine learning models become more embedded in critical applications, such as healthcare, finance, and autonomous systems, ensuring their trustworthiness is becoming increasingly essential. The course covers key topics including robustness against adversarial attacks, fairness in algorithmic decision-making, model interpretability, and privacy-preserving techniques such as differential privacy, federated learning and machine unlearning.",
+    happens: [
+      {
+        year: "2024 Spring",
+        link: "https://graph-and-geometric-learning.github.io/cpsc471-571-website-24spring",
+      },
+    ],
+  },
 ];
diff --git a/components/tag.tsx b/components/tag.tsx
index 8cef9a1..9b2971a 100644
--- a/components/tag.tsx
+++ b/components/tag.tsx
@@ -22,25 +22,25 @@ function PublicationTag({ tag }: { tag: Tag }) {
   switch (tag) {
     case Tag.Applications:
       name = "Applications";
-      color = "#ffe119";
+      color = "primary";
       break;
     case Tag.TrustworthyAI:
       name = "Trustworthy AI";
-      color = "#3cb44b";
+      color = "secondary";
       break;
     case Tag.MultiModalFoundationModel:
       name = "Multi-Modal Foundation Model";
-      color = "#4363d8";
+      color = "success";
       break;
     case Tag.GraphRepresentationLearning:
       name = "Graph Representation Learning";
-      color = "#f58231";
+      color = "warning";
       break;
   }
 
   return (
     <>
-      <Chip style={{ backgroundColor: color }}>{name}</Chip>
+      <Chip color={ color }>{name}</Chip>
     </>
   );
 }
diff --git a/config/people.ts b/config/people.ts
index 4392b8b..0229d6a 100644
--- a/config/people.ts
+++ b/config/people.ts
@@ -33,7 +33,7 @@ export const peopleList = {
       name: "Borui Wang",
       intro: "I am a final-year CS Ph.D. Student working on large language models, natural language processing and machine learning. My main research directions are in large language models, LLM-powered reinforcement learning, knowledge and logical reasoning, neuro-symbolic reasoning, graph neural networks, multimodal language grounding, and applications of LLMs and deep learning in quantitative finance and financial technology.",
       homepage: "https://borui-wang.github.io/",
-      photo: "people/borui.jpg",
+      photo: "/people/borui.jpg",
     },
     {
       name: "Tinglin Huang",
diff --git a/config/publications.ts b/config/publications.ts
index bab1531..8ef4eb2 100644
--- a/config/publications.ts
+++ b/config/publications.ts
@@ -16,16 +16,6 @@ export interface Publication {
 }
 
 export const publications = [
-  {
-    title: "HEART: Learning Better Representation of EHR Data with a Heterogeneous Relation-Aware Transformer",
-    authors: "T Huang, SA Rizvi, R Krishna Thakur, V Socrates, M Gupta, D Dijk, RA Taylor, R Ying",
-    venue: "Preprint",
-    code: null,
-    paper: "https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4904741",
-    abstract: "We propose HEART, a pretrained language model for structured EHR data. HEART seamlessly encodes heterogeneous medical entity information through a novel relation embedding module and a multi-level attention scheme.",
-    impact: "This novel pretrained framework, featuring a new architecture and dedicated objectives, can inspire future research on foundation models in EHR.",
-    tags: [Tag.Applications],
-  },
   {
     title: "Protein-Nucleic Acid Complex Modeling with Frame Averaging Transformer",
     authors: "Tinglin Huang, Zhenqiao Song, Rex Ying, Wengong Jin",
@@ -60,6 +50,16 @@ export const publications = [
     impact:
       "The learned high-order relationships achieve SOTA performance on predictive tasks and are demonstrated significantly more effective than traditional pairwise methods.",
   },
+  {
+    title: "HEART: Learning Better Representation of EHR Data with a Heterogeneous Relation-Aware Transformer",
+    authors: "T Huang, SA Rizvi, R Krishna Thakur, V Socrates, M Gupta, D Dijk, RA Taylor, R Ying",
+    venue: "Preprint",
+    code: null,
+    paper: "https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4904741",
+    abstract: "We propose HEART, a pretrained language model for structured EHR data. HEART seamlessly encodes heterogeneous medical entity information through a novel relation embedding module and a multi-level attention scheme.",
+    impact: "This novel pretrained framework, featuring a new architecture and dedicated objectives, can inspire future research on foundation models in EHR.",
+    tags: [Tag.Applications],
+  },
   {
     title: "Online Detection of Anomalies in Temporal Knowledge Graphs with Interpretability",
     authors: "Jiasheng Zhang, Rex Ying, Jie Shao",