4 ms·
Very interested to try this. I’ve built a number of quite complex prompts to do exactly this - cite from documents, with built-in safeguards to minimise halluc
by saaaaaam 2y ago
Very interested to try this.
I’ve built a number of quite complex prompts to do exactly this - cite from documents, with built-in safeguards to minimise hallucinations as far as possible.
That comes with a cost though - typically the output of one prompt is fed into another API call with a prompt that sense-checks/fact-checks the output against the source, and if there are problems it has to cycle back - with more API cost. We then human review a random selection of final outputs.
That works fine for non-critical applications but I’ve been cautious about rolling it out to chunkier problems.
Will start building with citations asap and see how it performs against what we already have. For me, Anthropic seems to be building stuff that has more meaningful application than what I’m seeing from Open AI - and by and large I’m finding Anthropic performs way way better for my use cases than Open AI - both via the API and the chatbot.
- diggan 2y ago> and by and large I’m finding Anthropic performs way way better for my use cases than Open AI - both via the API and the chatbot I also find Anthropic very useful as a whole, they seem to think a bit broader it feels like, compared to OpenAI. Question for curiosities sake, have you tried o1 "Pro Mode" before? It's a lot slower (can take minutes to reply) but been very good at "chunkier problems", if we understand that term similarly.
- saaaaaam 2y agoI have but only via the chat interface. I wasn’t particularly impressed, and for my purposes I’d rather use chained prompts via the API than try for “one shot”. However that could be because I’ve not amended my prompting style extensively enough. From what I’ve read o1 pro delivers most benefits from quite a different way of promoting.