-
Notifications
You must be signed in to change notification settings - Fork 0
/
index.html
98 lines (92 loc) · 3.33 KB
/
index.html
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
<!DOCTYPE html>
<html lang="en">
<head>
<meta charset="UTF-8">
<meta name="viewport" content="width=device-width, initial-scale=1.0">
<meta name="description" content="Embodied Meta Prompting: Introduction to a meta-prompting method for solving embodied tasks using LLM">
<meta name="keywords" content="Embodied AI, Meta Prompting, LLM, Subgoal, Expert, Object Detection">
<meta name="author" content="Yusuke Mikami">
<title>Embodied Meta Prompting</title>
<!-- SEO enhancement with structured data -->
<script type="application/ld+json">
{
"@context": "https://schema.org",
"@type": "WebPage",
"name": "Embodied Meta Prompting",
"description": "A meta-prompting method for tackling embodied tasks step by step. This approach uses LLMs to generate experts for each subgoal, aiming for efficient task completion.",
"author": {
"@type": "Person",
"name": "Yusuke Mikami"
}
}
</script>
<style>
body {
font-family: Arial, sans-serif;
line-height: 1.6;
margin: 0;
padding: 0;
background-color: #f4f4f4;
color: #333;
}
header {
background-color: #333;
color: #fff;
padding: 1rem 0;
text-align: center;
}
h1 {
font-size: 2.5rem;
margin-bottom: 1rem;
color: #fff; /* タイトルの文字色を白に変更 */
}
.container {
max-width: 960px;
margin: 2rem auto;
padding: 1rem;
background-color: #fff;
box-shadow: 0 0 10px rgba(0, 0, 0, 0.1);
}
h2 {
color: #333;
}
p {
margin-bottom: 1rem;
line-height: 1.7;
}
footer {
text-align: center;
padding: 2rem 0;
background-color: #333;
color: #fff;
margin-top: 2rem;
}
a {
color: #fff;
text-decoration: none;
}
</style>
</head>
<body>
<header>
<h1>Embodied Meta Prompting</h1>
</header>
<div class="container">
<section>
<h2>Project Overview</h2>
<p>"Embodied Meta Prompting" is a project that emphasizes the importance of solving embodied tasks step by step. This method leverages Large Language Models (LLMs) to generate experts for each subgoal, aiming to effectively accomplish tasks.</p>
</section>
<section>
<h2>Meta-Prompting Method</h2>
<p>Through the meta-prompting method, multiple Vision Language Models (VLMs) can function as specialists for different sub-tasks. For example, an expert specialized in object detection can be generated and coordinated with other specialists working on different subgoals.</p>
</section>
<section>
<h2>Project Goals</h2>
<p>The ultimate goal of this project is to establish a new method for efficiently and effectively solving complex embodied tasks. By using LLMs, a meta-approach to task-solving becomes possible, enabling the integration of multiple expert models.</p>
</section>
</div>
<footer>
<p>© 2024 Yusuke Mikami - <a href="https://github.com/shure-dev/embodied-meta-prompting" target="_blank">GitHub Project Page</a></p>
</footer>
</body>
</html>