thibaultamartin, to writing
@thibaultamartin@mamot.fr avatar

I am not a native English speaker. I make grammar mistakes, and this can make me difficult to understand. Instead of nagging colleagues to proofread my writing, I decided to find a tool to correct my mistakes and help me learn along the way.

I evaluated four popular tools to make an informed decision. Here is a tale of Privacy Policies, slapping OpenAI on products to catch up with competition, and grammar checkers.

https://ergaster.org/posts/2024/02/26-writing-is-hard/

Colarusso, to ai
@Colarusso@mastodon.social avatar

I find this project, based on the LegalBench dataset,¹ which aims to benchmark LLMs on legal reasoning tasks interesting.

https://www.vals.ai/

TL:DR: Curious about which language models perform best on legal reasoning tasks? The latest evaluation reveals that Open AI's GPT-4 takes the lead, followed closely by Google's Gemini Pro.

¹ https://hazyresearch.stanford.edu/legalbench/

bwinbwin, to ai

instead of called them AIs, can we call them what they really are...LMs (language models)?

AI just over-inflates their capabilities and leads to poor understanding among non-experts.

jd7h, to LLMs
@jd7h@fosstodon.org avatar

"Stochastic parrot" was included in the shortlist for Word of the Year 2023!

https://americandialect.org/wp-content/uploads/2024/01/2023-Word-of-the-Year-PRESS-RELEASE.pdf

ocramz, (edited ) to random
@ocramz@sigmoid.social avatar

I've been thinking about program understanding, and about how to encourage to do compositional/verifiable on program text ("statically").

This is my latest work around this: https://aclanthology.org/2023.findings-emnlp.601/

If, as some recent literature suggest, transformer-based LMs are not more expressive than regexps, this line of thinking is doomed, but at least it could be a valuable heuristic and complementary to rigorous .

Everybody’s talking about Mistral, an upstart French challenger to OpenAI (arstechnica.com)

On Monday, Mistral AI announced a new AI language model called Mixtral 8x7B, a "mixture of experts" (MoE) model with open weights that reportedly truly matches OpenAI's GPT-3.5 in performance—an achievement that has been claimed by others in the past but is being taken seriously by AI heavyweights such as OpenAI's Andrej...

lampinen, to ai

Very excited to share a substantially updated version of our preprint “Language models show human-like content effects on reasoning tasks!” TL;DR: LMs and humans show strikingly similar patterns in how the content of a logic problem affects their answers. Thread: 1/10

nextcloud, to ai
@nextcloud@mastodon.xyz avatar

😲Did you know compliments can influence large-language models?

They're even open to Argentinian Spanish & Swiss German slang.

Discover how Nextcloud AI is tackling ethical concerns differently:

https://nextcloud.com/blog/nextcloud-ethical-ai-rating/

lysander07, to llm

Many new and interesting topics in our upcoming - Foundations and Applications online lecture at

lysander07, to fediverse

Finally, The Time Traveler’s Guide to Research: Analyzing Fictitious Research Themes in the ESWC „Next 20 Years“ Track has been published by Heiko Paulheim (still not present in the ) and Irene Celino https://arxiv.org/abs/2309.13939

mjgardner, to ArtificialIntelligence

I keep thinking of this bit from as people keep adding “” to stuff: https://www.youtube.com/watch?v=kAqIJZeeXEc

“ has what minds crave! It has !”

“"

lysander07, to ArtificialIntelligence

Next step in our brief timeline of (large) from our lecture was statistical language modeling with n-grams based on large text corpora as introduced and popularized by Frederick Jelinek and Stanley F. Chen using statistical tricks like Bayes Theorem, Markov Assumption, and Maximum Likelihood Estimation, etc.
Slides: https://drive.google.com/file/d/1atNvMYNkeKDwXP3olHXzloa09S5pzjXb/view?usp=drive_link
@fizise

jd7h, to LLMs
@jd7h@fosstodon.org avatar

I'm taking some time today to test a few new libraries/tools.
These CLI tools for working with llms by @simon work like a charm! And they support unix pipes. <3

More info here: https://llm.datasette.io/en/latest/index.html

rysiek, to ai
@rysiek@mstdn.social avatar

Dear , there's been some buzz recently about that are not gigantic black boxes, and in general, developed as .

There's this Google internal document, for example, that points out FLOSS community is close to eating Google's and OpenAI's cake:
ttps://www.semianalysis.com/p/google-we-have-no-moat-and-neither

So here is my question to you:

What are the best examples of useful, small, on-device models already out there?

:boost_requested:

  • All
  • Subscribed
  • Moderated
  • Favorites
  • •
  • megavids
  • kavyap
  • DreamBathrooms
  • thenastyranch
  • magazineikmin
  • osvaldo12
  • ethstaker
  • Youngstown
  • mdbf
  • slotface
  • rosin
  • everett
  • ngwrru68w68
  • khanakhh
  • JUstTest
  • InstantRegret
  • GTA5RPClips
  • Durango
  • normalnudes
  • cubers
  • tacticalgear
  • cisconetworking
  • tester
  • modclub
  • provamag3
  • anitta
  • Leos
  • lostlight
  • All magazines
    •