1<html> 2<head> 3 <meta charset="utf-8"> 4 <title>Can ChatGPT Detect Intent? Evaluating Large Language Models for Spoken Language Understanding</title>
vendor: 65 bytes, lines 4-5
4 5 <script async src="https://www.googletagmanager.com/gtag/js?id=
5G-3T2S9NK8XL
vendor: 16 bytes, lines 5-6
5"></script> 6
6<script> 7
vendor: 177 bytes, lines 7-15
7window.dataLayer = window.dataLayer || []; 8 9 function gtag() { 10 dataLayer.push(arguments); 11 } 12 13 gtag('js', new Date()); 14 15 gtag('config', '
15G-3T2S9NK8XL
vendor: 8 bytes, lines 15-16
15'); 16
16</script>
16 17</head> 18<body> 19<h1>Can ChatGPT Detect Intent? Evaluating Large Language Models for Spoken Language Understanding</h1> 20<p> 21Here we provide examples of the prompts used in the evaluation. For the cases with fewer N<30 shots, simply keep only the first N examples. 22<ul> 23 <li>Prompts for ChatGPT: SLURP Intent classification <a href="chatgpt_slurp.txt">[0-shot]</a>, <a href="chatgpt_slurp.30_shot.txt">[30-shot]</a>, <a href="chatgpt_slurp.20_shot.xlong.txt">[20-shot, verbose]</a>, <a href="chatgpt_minds14.ko.14_shot.txt">[MINDS-14 Korean 14-shot]</a>, SLURP slot filling <a href="chatgpt_slurp_ent.30_shot.des.txt">[30-shot]</a></li> 24 <li>Prompts for GPT3.5: SLURP Intent classification <a href="gpt3_slurp.30_shot.txt">[30-shot]</a>, <a href="gpt3_slurp.5_shot.asr.noise.chain.txt">[ASR with chain-of-thought]</a></li> 25 <li>Prompts for other autoregressive language models: SLURP Intent classification <a href="arlm_slurp.30_shot.txt">[30-shot]</a></li> 26</ul> 27</body> 28</html>
Line numbers count LF bytes from the start of the resource, as the search results do. Vendor segments are library code the classifier recognised; they are stored but not indexed. Bytes are shown as Latin1 characters, one per byte.