TV Feeder

Supercomputer telecommunications company. @TVFeederInc @JasonPGilbert87

[TV Feeder @WatchTVFeeder]

https://t.co/mRM0e4V5TQ

[Devmos 2024 @devmos2024]

Babylon - Bitcoin Staking

#Osmocon2023

[Open Source Economy @OS_Economy]

Retweet if you agree!

[TV Feeder @WatchTVFeeder]

Panel Discussion- Reimagine Media & Content Ecosystem with Web3 & AI https://t.co/rxzSLwLSkl via @Blockcast.cc- News on Blockchain, DLT, Cryptocurrency

[TV Feeder @WatchTVFeeder]

@ElonMuskAOC @p2pdht Absolutely nothing because inflation would also go up 1M percent.

[TV Feeder @WatchTVFeeder]

https://t.co/sTDzxkEiuR

[Vergil ✺ @madebyvergil]

100% Multimodal LLMs will be the new OS.

It's the only way we can have lighter devices (Laptops, XR Glasses, BCIs), with bigger batteries, less heat, non built-in obsolescence.

Less demand for hardware -> Cheaper materials -> More access to powerful computing for the masses.

[Shane McGrath @bloodbatmcgrath]

When people talk about multimodal, what they really mean is brain activity to speech.

https://t.co/NVL8d1bcev

[Tricky @trickylabyrinth]

someone explain how this's working? what is actually making llama multimodal?

[Joseph Nelson @josephofiowa]

this is the best breakdown on the state of multimodality that i've read, esp why CLIP has been so key in getting us here

https://t.co/1lL1d88WXF by @chipro

[ml-sanity bot @arxivsanitybot]

https://t.co/WSLYIgN0qj A new study has shown that simple modifications to a multimodal model can achieve state-of-the-art results on various benchmarks with minimal data usage and training time. Code and model will be publicly available.

[ml-sanity bot @arxivsanitybot]

https://t.co/FM4xZaux97 Researchers propose a new two-stage pruning approach for Large Vision-Language Models that achieves significant performance improvements on various multimodal and unimodal datasets.

[Fum（ファム）🇬🇧 AIエージェント研究 @Fumweb3]

#AI #LLMs
こっちは、Rekaが発表したYasa-1というマルチモーダルなLLMだけど、インプットとして、画像やオーディオ、そして短い動画にも対応してるらしい🤩
https://t.co/xiOZ5FLeHS

まだプライベート・プレビュー段階で試せてないけど、デモでは、無音の動画の内容も認識でき、さらにその次に何が起こるかを予測していたり、かみなや雨の音を認識して説明してたりしてた😳

20言語サポート、検索エンジン対応、さらにコード・インタープリターと、かなりモリモリな内容に��っているみたいなので、言語性能次第だけど、期待できそう✨

[AK @_akhaliq]

Improved Baselines with Visual Instruction Tuning

paper page: https://t.co/YEabbP3Ic3

Large multimodal models (LMM) have recently shown encouraging progress with visual instruction tuning. In this note, we show that the fully-connected vision-language cross-modal connector in LLaVA is surprisingly powerful and data-efficient. With simple modifications to LLaVA, namely, using CLIP-ViT-L-336px with an MLP projection and adding academic-task-oriented VQA data with simple response formatting prompts, we establish stronger baselines that achieve state-of-the-art across 11 benchmarks. Our final 13B checkpoint uses merely 1.2M publicly available data, and finishes full training in ~1 day on a single 8-A100 node. We hope this can make state-of-the-art LMM research more accessible.

[Patrick's AIBuzzNews @AIBuzzNews]

LLaVA can be brutally honest and even insulting.

There are no restrictions or filters to hold it back.

LLaVA is a multimodal LLM on @huggingface like ChatGPT will be soon.

I asked it to criticize and roast my look, and I do not regret it.😅

Check ALT for the link and prompt.

[Brian Roemmele @BrianRoemmele]

ChatGPT-4V Multimodal make a great effort to present the IBM System 370 control board function from just a single image.

More questions produced a use for each button, switch or dial.

[👀 @JoyaKhanzeb]

Latest Advancements in the Field of Multimodal AI: (ChatGPT + DALLE 3) + (Google BARD + Extensions) and many more….

[Brendan Paul Jacobs @BrendanPJacobs]

Today I was pleased to introduce Provisional Multimodal Research at the FabLearn/Constructionism 2023 conference at Columbia University. #provisionalmultimodality

[Zihao He @ZihaoHe95]

One paper on multimodal (audio-text) fusion got accepted by #EMNLP2023 ! Stay tuned for more details!