cmpatino HF Staff commited on
Commit
ac128a2
·
verified ·
1 Parent(s): 7dbe2c3

Text: end the rules sentence at the parenthesis

Browse files
Files changed (1) hide show
  1. site/index.html +1 -1
site/index.html CHANGED
@@ -52,7 +52,7 @@
52
  <p>LLMs drive today's chatbots and agents, but conversational programs are much older than LLMs. Sixty years ago, <a href="https://en.wikipedia.org/wiki/ELIZA" target="_blank" rel="noopener">ELIZA</a> held
53
  conversations by matching patterns in what the user typed and answering with predefined templates.</p>
54
  <p>We wanted to revisit that idea at a much larger scale, so we gave a swarm of LLM agents the mission of building a chatbot out of
55
- plain Python rules (if-else branches, regex patterns and lookup tables), with no model inside. The agents wrote and evaluated over
56
  5,000 versions of the bot, and its win rate on <a href="https://huggingface.co/datasets/HuggingFaceSwarms/SmolTalkBench" target="_blank" rel="noopener">SmolTalkBench</a>, where an LLM judge compares each reply with a reference answer,
57
  climbed from almost 0% to 26%.</p>
58
  <p>The result is PyBot: about 470k lines of Python that answer everyday questions, plus 550k lines of tests the agents wrote to keep
 
52
  <p>LLMs drive today's chatbots and agents, but conversational programs are much older than LLMs. Sixty years ago, <a href="https://en.wikipedia.org/wiki/ELIZA" target="_blank" rel="noopener">ELIZA</a> held
53
  conversations by matching patterns in what the user typed and answering with predefined templates.</p>
54
  <p>We wanted to revisit that idea at a much larger scale, so we gave a swarm of LLM agents the mission of building a chatbot out of
55
+ plain Python rules (if-else branches, regex patterns and lookup tables). The agents wrote and evaluated over
56
  5,000 versions of the bot, and its win rate on <a href="https://huggingface.co/datasets/HuggingFaceSwarms/SmolTalkBench" target="_blank" rel="noopener">SmolTalkBench</a>, where an LLM judge compares each reply with a reference answer,
57
  climbed from almost 0% to 26%.</p>
58
  <p>The result is PyBot: about 470k lines of Python that answer everyday questions, plus 550k lines of tests the agents wrote to keep