Загрузка видео...

Не удалось загрузить видео

На главную

MolmoPoint Better Pointing for VLMs with Grounding Tokens paper: models: app:

_akhaliq's profile picture

AK

506,744 subscribers

20,057 просмотров • 5 месяцев назад •via X (Twitter)

LIVE

Anya Rossi• Live Now

Private livecam show

Комментарии: 0

Нет доступных комментариев

Здесь появятся комментарии из оригинального поста

Похожие видео

Grounding World Simulation Models in a Real-World Metropolis paper:
2:04

Grounding World Simulation Models in a Real-World Metropolis paper:

AK

33,812 просмотров • 5 месяцев назад

Vision-language models (VLMs) can see well, but they struggle to reason. In this episode, Antonia Wüst (PhD researcher, TU Darmstadt) explains how combining VLMs with program synthesis yields more reliable visual reasoning, with fewer tokens than chain-of-thought.
0:46

Vision-language models (VLMs) can see well, but they struggle to reason. In this episode, Antonia Wüst (PhD researcher, TU Darmstadt) explains how combining VLMs with program synthesis yields more reliable visual reasoning, with fewer tokens than chain-of-thought.

Ndea

22,130 просмотров • 7 месяцев назад

Excited to share our new work: StreamingVLM! 🚀 We tackle a major challenge for Vision-Language Models (VLMs): understanding infinite video streams in real-time without latency blowing up or running out of memory. Paper: Code:
2:26

Excited to share our new work: StreamingVLM! 🚀 We tackle a major challenge for Vision-Language Models (VLMs): understanding infinite video streams in real-time without latency blowing up or running out of memory. Paper: Code:

Guangxuan Xiao

92,141 просмотров • 10 месяцев назад

New for AI app devs: Billing for LLM tokens with Stripe - Charge users for token usage (+X% margin) - Call LLMs with our proxy API or via partners: , @cloudflare, Vercel, Helicone - Update prices as models change costs Get access here:
2:04

New for AI app devs: Billing for LLM tokens with Stripe - Charge users for token usage (+X% margin) - Call LLMs with our proxy API or via partners: , @cloudflare, Vercel, Helicone - Update prices as models change costs Get access here:

Jeff Weinstein

41,723 просмотров • 10 месяцев назад

(1/n) Time to unify your favorite visual generative models, VLMs, and simulators for controllable visual generation—Introducing a Product of Experts (PoE) framework for inference-time knowledge composition from heterogeneous models.
0:55

(1/n) Time to unify your favorite visual generative models, VLMs, and simulators for controllable visual generation—Introducing a Product of Experts (PoE) framework for inference-time knowledge composition from heterogeneous models.

Yunzhi Zhang

49,090 просмотров • 1 год назад

⚠️SHOCKING BLOOD TEST RESULTS AFTER GROUNDING❗️ “Blood 🩸 cells get freed up for better circulation in the body after grounding.” - Dr. Cimone Kamel.
1:14

⚠️SHOCKING BLOOD TEST RESULTS AFTER GROUNDING❗️ “Blood 🩸 cells get freed up for better circulation in the body after grounding.” - Dr. Cimone Kamel.

𝐏𝐇𝐎𝐄𝐍𝐈𝐗🐦‍🔥𝐃𝐑𝐀𝐆𝐎𝐍

37,560 просмотров • 1 год назад

Punchline: World models == VQA (about the future)! Planning with world models can be powerful for robotics/control. But most world models are video generators trained to predict everything, including irrelevant pixels and distractions. We ask - what if a world model only predicted the semantic information necessary for decision-making? Introducing Semantic World Models (SWM). Given an observation and an action sequence, SWMs cast modeling as answering textual questions about the future outcome resulting from the actions. Recasting world modeling as a VQA problem lets us directly leverage the pretrained knowledge and machinery of VLMs for generalizable modeling. We had a lot of fun thinking about how this work helps connect these two seemingly very different fields of study - VLMs and world models! 🧵(1/6) Paper: Fun demo:
0:49

Punchline: World models == VQA (about the future)! Planning with world models can be powerful for robotics/control. But most world models are video generators trained to predict everything, including irrelevant pixels and distractions. We ask - what if a world model only predicted the semantic information necessary for decision-making? Introducing Semantic World Models (SWM). Given an observation and an action sequence, SWMs cast modeling as answering textual questions about the future outcome resulting from the actions. Recasting world modeling as a VQA problem lets us directly leverage the pretrained knowledge and machinery of VLMs for generalizable modeling. We had a lot of fun thinking about how this work helps connect these two seemingly very different fields of study - VLMs and world models! 🧵(1/6) Paper: Fun demo:

Abhishek Gupta

61,390 просмотров • 10 месяцев назад

👀Humans compare images by looking back and forth. Many open-weight VLMs encode each image independently, and defer comparison to the LM. We introduce SVE: Stateful Visual Encoders for Vision-Language Models, where the visual encoder itself becomes change-aware. 🌐Project: 📰Paper: 💻Code: 1/n
1:11

👀Humans compare images by looking back and forth. Many open-weight VLMs encode each image independently, and defer comparison to the LM. We introduce SVE: Stateful Visual Encoders for Vision-Language Models, where the visual encoder itself becomes change-aware. 🌐Project: 📰Paper: 💻Code: 1/n

Zirui "Colin" Wang

63,292 просмотров • 2 месяцев назад

🤯ByteDance just Open Sourced UI-TARS - 2 SOTA models (7B & 72B) + a PC/MacOS app to control your computer with vLMS And they are not messing around, beating GPT-4o and Claude, SOTA across 10 benchmarks Will you be installing this on your pc?
3:55

🤯ByteDance just Open Sourced UI-TARS - 2 SOTA models (7B & 72B) + a PC/MacOS app to control your computer with vLMS And they are not messing around, beating GPT-4o and Claude, SOTA across 10 benchmarks Will you be installing this on your pc?

Alex Volkov

69,738 просмотров • 1 год назад

🚨In our NeurIPS paper, we bring encoder-decoders back.. for diffusion language models! ⚡️Encoder-decoders make diffusion sampling fast: a small (fast) decoder denoises tokens progressively and a large (slower) encoder represents clean context.
0:31

🚨In our NeurIPS paper, we bring encoder-decoders back.. for diffusion language models! ⚡️Encoder-decoders make diffusion sampling fast: a small (fast) decoder denoises tokens progressively and a large (slower) encoder represents clean context.

Marianne Arriola @ICML

32,073 просмотров • 10 месяцев назад

I asked for a 3D spinning 'G' Google logo. Gemini 3 Flash was 3x faster, used 50% fewer tokens and gave much better results. Gemini 3 Flash: 6.6k tokens, 38s Gemini 2.5 Pro: 13k tokens, 108s You can use this AI Studio app to do your own comparisons:
2:04

I asked for a 3D spinning 'G' Google logo. Gemini 3 Flash was 3x faster, used 50% fewer tokens and gave much better results. Gemini 3 Flash: 6.6k tokens, 38s Gemini 2.5 Pro: 13k tokens, 108s You can use this AI Studio app to do your own comparisons:

fofr

46,503 просмотров • 8 месяцев назад

MiniCPM-V 4.5 achieves an average score of 77.0 on OpenCompass, a comprehensive evaluation of 8 popular benchmarks. With only 8B parameters, it surpasses widely used proprietary models like GPT-4o-latest, Gemini-2.0 Pro, and strong open-source models like Qwen2.5-VL 72B powered by a new unified 3D-Resampler over images and videos, MiniCPM-V 4.5 can now achieve 96x compression rate for video tokens, where 6 448x448 video frames can be jointly compressed into 64 video tokens (normally 1,536 tokens for most MLLMs). vibe coding a Video Chat AI app with MiniCPM-V-4.5 in anycoder
1:32

MiniCPM-V 4.5 achieves an average score of 77.0 on OpenCompass, a comprehensive evaluation of 8 popular benchmarks. With only 8B parameters, it surpasses widely used proprietary models like GPT-4o-latest, Gemini-2.0 Pro, and strong open-source models like Qwen2.5-VL 72B powered by a new unified 3D-Resampler over images and videos, MiniCPM-V 4.5 can now achieve 96x compression rate for video tokens, where 6 448x448 video frames can be jointly compressed into 64 video tokens (normally 1,536 tokens for most MLLMs). vibe coding a Video Chat AI app with MiniCPM-V-4.5 in anycoder

AK

19,088 просмотров • 11 месяцев назад

No blind dates. Just verified tokens. Now live on Base App and Farcaster : the dating mini-app for tokens. Profiles > vibes > swipe. Right = HOT, left = NOT. Meet your onchain type👇
0:50

No blind dates. Just verified tokens. Now live on Base App and Farcaster : the dating mini-app for tokens. Profiles > vibes > swipe. Right = HOT, left = NOT. Meet your onchain type👇

WachAI

66,433 просмотров • 1 год назад

We're introducing new stereo models for MusicGen. By extending the delay codebook pattern to cover tokens from both left & right channels, these models can generate stereo output with no extra computational cost vs previous models. Try MusicGen on 🤗 ➡️
0:33

We're introducing new stereo models for MusicGen. By extending the delay codebook pattern to cover tokens from both left & right channels, these models can generate stereo output with no extra computational cost vs previous models. Try MusicGen on 🤗 ➡️

AI at Meta

214,803 просмотров • 2 лет назад

Gold-backed tokens shouldn’t rely on trust. That’s why Jiritsu partnered with Brinks, verifying gold at the vault- not just on paper.
1:23

Gold-backed tokens shouldn’t rely on trust. That’s why Jiritsu partnered with Brinks, verifying gold at the vault- not just on paper.

Jiritsu Network

62,767 просмотров • 1 год назад

meet the app gallery a curated library of vibe coded projects built with powerful AI integrations like Nano Banana, Maps Grounding and more remix any app with a single prompt and continue building to make it your own
0:29

meet the app gallery a curated library of vibe coded projects built with powerful AI integrations like Nano Banana, Maps Grounding and more remix any app with a single prompt and continue building to make it your own

Google AI Studio

74,784 просмотров • 10 месяцев назад

Chat with Horny Cam Models on Superlive App😍📷
0:27

Chat with Horny Cam Models on Superlive App😍📷

Daily P!MP💦

17,294 просмотров • 1 месяц назад