Se connecter

Recherches récentes

Aucune recherche récente

Options de recherche

Disponible uniquement lorsque vous êtes connecté.
framapiaf.org est l'un des nombreux serveurs Mastodon indépendants que vous pouvez utiliser pour participer au fédiverse.
Un service Mastodon fourni par l'association d’éducation populaire Framasoft.

Administré par :

Statistiques du serveur :

1,4K
comptes actifs

framapiaf.org: À proposÉtatAnnuaire des profilsPolitique de confidentialité

Mastodon: À proposTélécharger l’applicationRaccourcis clavierVoir le code sourcev4.3.7

#llm

508 messages200 participants38 messages aujourd’hui
Public
Edizioni Novilunio

Negli ultimi due anni vari settori sono stati messi in difficoltà dall'arrivo destabilizzante di #LLM, #IA e altre tecnologie generative. Il settore editoriale, basandosi sulle parole che tali strumenti generano con più semplicità, non ne è stato immune.

Noi di Novilunio abbiamo riflettuto su cosa significa generare #arte, fino a che punto e a che costo possa essere ancora considerata tale.

(La prima metà dell'articolo è una traduzione di un post di @pluralistic)

[edizioninovilunio.it/perche-no]

Edizioni Novilunio · Perché non mi piace l’arte generativa - Edizioni NovilunioPerché non ci piace l'arte generativa, e possiamo davvero considerarla tale? Riprendendo un articolo di Cory Doctorow, diamo la nostra opinione.
Public
FOSSASIA

👁️‍ Learn how to spot and visualize LLM hallucinations — gain clarity on when AI models go off track and how to better evaluate their outputs with Anand Subramanian at #FOSSASIASummit2025

🔗 Click here youtu.be/FEz3rMwArQs to watch on the FOSSASIA YouTube channel.

youtu.be- YouTubeProfitez des vidéos et de la musique que vous aimez, mettez en ligne des contenus originaux, et partagez-les avec vos amis, vos proches et le monde entier.
#LLM#AI#FOSSASIA
Public
AlexCrimi

I have seen that m5stack has made an #LLM module for the M3core model. Do you think it is worth it?
The llm module is about 60 USD and the M3core is 80 USD. I thought that with money you buy the most advanced #raspberrypi and you put a display, no?

Public
Otavio R. Piske

I had a lot of fun recently dealing with function calls and it's a good read despite causing me a bit of a panic attack.

"... While you must rely on the model to decide when and how to use tools, it may not always align with your expectations, despite your best efforts ..."

queue.acm.org/detail.cfm?ref=r

Granted: a good of chunk of the problems was caused by bugs I introduced myself, but having the #LLM in the middle didn't make things easy to deal with.

queue.acm.orgFrom Function Frustrations to Framework Flexibility - ACM Queue
Public
ndocist

An automated framework for assessing how well LLMs cite relevant medical references | Nature Communications
"50% to 90% of seven popular #LLM responses are not fully supported, and sometimes contradicted, by the sources they cite" GPT-4o, Claude v2.1, Mistral Medium and Gemini (LLM, RAG or API). Moins d'hallucinations avec les RAG
nature.com/articles/s41467-025

NatureAn automated framework for assessing how well LLMs cite relevant medical references - Nature CommunicationsAssessing the degree to which medical large language models reliably convey existing, trustworthy knowledge is crucial. This study introduces SourceCheckup, an automated framework revealing that large language models frequently cite medical references that do not fully support, or even contradict, their responses, showing significant gaps in reliability for clinical use.
ExplorerFlux en direct
À propos