Live data from Hacker News

Mercury 2.5

inceptionlabs.ai

31–40 of 58 posts

Re: Mercury 2.5

#31

Oh great a new model annoncemzzzzzz ZZZZZZZZZ

Not just a model - It's a diffusion model. Instead of Next Token prediction it builds the entire page at once and then denoises it. Kind of mind blowing when you watch it happen.

Ok well that is actually interesting, thank you. Instead of OpenAI beta 6.3pre3 “fairyfloss”

Re: Mercury 2.5

#32
I’m using this model to „rerank” results from vector store. The model is provided a set of results and asked to produce a string of 1s and 0s where the offset reflects the position in the result set. The prompt goes along the line „do this set of result match the provided query X”. Works like a charm, normally I would use a small non reasoning model, but given how Mercury produces the output its blazingly fast - which is what I was optimizing for - not to increase the search latency. It helped improving our search in a way that reranker could get close to.

Re: Mercury 2.5

#34
I tried at it creative writing - and, with thinking off, it was considerably better than Mercury 2 and generally good in fact, not very sloppy. Now with thinking on, it got worse, began hallucinating things; this is something I've noticed with all recent models - enabling reasoning causes hallucinations in creative writing assignments.

Re: Mercury 2.5

#35

I tried at it creative writing - and, with thinking off, it was considerably better than Mercury 2 and generally good in fact, not very sloppy. Now with thinking on, it got worse, began hallucinating things; this is something I've noticed with all recent models - enabling reasoning causes hallucinations in creative writing assignments.

Similar, I've found that on most recent models I've tried, reasoning will cause it to spend quite a lot of time thinking about whether even the most wholesome fictional scenario is something it's allowed to write about, which in turn causes it to start hallucinating details. For creative writing I've found the minimum of thinking is the best way.

Re: Mercury 2.5

#36

Oh great a new model annoncemzzzzzz ZZZZZZZZZ

I think we should be happy when this happens as it encourages more players to be active in AI, and us not rely on only two companies

Re: Mercury 2.5

#38

I tried at it creative writing - and, with thinking off, it was considerably better than Mercury 2 and generally good in fact, not very sloppy. Now with thinking on, it got worse, began hallucinating things; this is something I've noticed with all recent models - enabling reasoning causes hallucinations in creative writing assignments.

Similar, I've found that on most recent models I've tried, reasoning will cause it to spend quite a lot of time thinking about whether even the most wholesome fictional scenario is something it's allowed to write about, which in turn causes it to start hallucinating details. For creative writing I've found the minimum of thinking is the best way.

i’ve found this for myself as well, when its time for me to write creatively

Re: Mercury 2.5

#39

I tried at it creative writing - and, with thinking off, it was considerably better than Mercury 2 and generally good in fact, not very sloppy. Now with thinking on, it got worse, began hallucinating things; this is something I've noticed with all recent models - enabling reasoning causes hallucinations in creative writing assignments.

Could you explain how you define hallucination in the context of creativity? Lack of consistency aside, the two look quite similar to me - or would you consider the existence of orcs and elves in LOTR to be hallucinations, too?

Re: Mercury 2.5

#40
post #39

I tried at it creative writing - and, with thinking off, it was considerably better than Mercury 2 and generally good in fact, not very sloppy. Now with thinking on, it got worse, began hallucinating things; this is something I've noticed with all recent models - enabling reasoning causes hallucinations in creative writing assignments.

Could you explain how you define hallucination in the context of creativity? Lack of consistency aside, the two look quite similar to me - or would you consider the existence of orcs and elves in LOTR to be hallucinations, too?

> Could you explain how you define hallucination in the context of creativity

Mostly as you've mentioned, "lack of consistency", which manifests in variety of ways - presence of cellphones in historical settings (I'd call them "global inconsistency") and a character for whatever reason suddenly smoking a cigarette ("local inconsistency") I've never mentioned nor present in the context.

Post reply on HN