Re:LM2: Large Memory Models
by ziofill on Thursday February 13, 2025 @07:03PM
(#43043056)
Missed opportunity to call them LMM and enjoy the onslaught of typos
[ Parent
| on HN ]
They're still language models
LMLM
Re:LM2: Large Memory Models
by stuartjohnson12 on Thursday February 13, 2025 @07:30PM
(#43043222)
I can't wait to enhance my STT-RAG-LLM-TTS with some LMMs
Re:LM2: Large Memory Models
by bloomingkales on Thursday February 13, 2025 @08:32PM
(#43043660)
You forgot SOTA, that one is flying around a lot.
Re:LM2: Large Memory Models
by janderson215 on Thursday February 13, 2025 @09:58PM
(#43044289)
Also left out SoDoSoPa
Re:LM2: Large Memory Models
by pylotlight on Thursday February 13, 2025 @08:58PM
(#43043854)
Wouldn't it be STT-LLM-RAG-TTS
That's why they named it LM2 ..
Re:LM2: Large Memory Models
by pylotlight on Thursday February 13, 2025 @08:53PM
(#43043823)
Seems like LM² would have made more sense in that case.
Or L²M²
Or LLMM
(LM)²
Re:LM2: Large Memory Models
by DiogenesKynikos on Friday February 14, 2025 @09:04AM
(#43048375)
The question is whether L and M commute.
Because that makes it harder to type out
Or Models with Large Memory. MLMs :)
Re:LM2: Large Memory Models
by x86hacker1010 on Friday February 14, 2025 @12:36AM
(#43045216)
This seems to be the most appropriate
Re:LM2: Large Memory Models
by DebtDeflation on Friday February 14, 2025 @07:35AM
(#43047738)
Just as Graph RAG should have just been GAG.
Some Chinese researchers have worked on copper nanotubes and abbreviated them as CuNTs [0].
[0]: https://www.researchgate.net/publication/260800015_Structura...