Muse Spark Open delivers 816B params, native multimodal power on Macs with 1 TB RAM. #AI #OpenWeight #MuseSpark #Mac #Meta #Multimodal thedailytechfeed.com/metas-muse-s...

🤖 Apertus 1.5 Enhances Reasoning with Adaptive Thinking Mode

Apertus 1.5 adds images, audio, a thinking mode and a fourfold longer context window, on top of the open weights and open data that defined the previous generation. The...

#InferenceOptimization #Multimodal #LLM #AI #AIPulse

🤖 Apple's Tri-Modal Diffusion Model Hits 6.4T Tokens, Raises Evaluation Questions

The study is the largest systematic open survey of multimodal discrete diffusion models to date, pretraining from scratch on text,...

#ModelTraining #Multimodal #BenchmarksEvaluation #AI #AIPulse

トリモーダルマスク拡散モデルの設計空間：大規模マルチモーダル事前学習の全貌

テキスト・画像・音声の3モダリティを統合したマスク拡散モデルの設計空間とスケーリング則に関する研究レポート。

#DiffusionModels #Multimodal #ScalingLaws #MachineLearning

New #Expert Certification:

Mechanisms of Multimodal Synchronization: Insights from Decoder-Based Video-Text-to-Speech Synthesis

Akshita Gupta, Tatiana Likhomanenko, Karren Dai Yang et al.

https://openreview.net/forum?id=D5QpMy6yUH

#multimodal #encoding #synchronization

🤖 Google's Argon Model Leads in Embodied Reasoning Benchmark

Gemini 4 Argon scores 87 percent on the Embodied Reasoning benchmark, which tests physical reasoning in 3D environments. The score beats every other frontier...

#BenchmarksEvaluation #Multimodal #Robotics #AI #AIPulse

Gemini 4 Argon for Robotics：物理的AGIを加速させるマルチモーダル汎用推論モデルの全貌

物理的AGIの実現に向けた次世代ロボット制御モデル「Gemini 4 Argon」のアーキテクチャと機能詳解。

#Gemini #Robotics #MuJoCo #VLA #Multimodal

🤖 Z.ai's GLM-5.3 Model Matches Frontier AI Performance at Lower Cost

The headline number is a benchmark score: 19.17 per cent pass rate on Harvey's Legal Agent Benchmark, against 6.67 per cent for the base model and roughly three...

#BenchmarksEvaluation #Multimodal #LLM #AI #AIPulse

New arXiv paper UkrMeme introduces a 200-meme Ukrainian wartime dataset and benchmarks vision-language models for persuasion technique detection using zero-shot prompting. Worth a look for anyone testing multimodal VLM prompting…

#AI #NLP #Prompting #Multimodal
https://arxiv.org/abs/2610.13783

🤖 Ai2 Unveils Next Olmo Model with Hybrid Architecture

The next model is built on the finding that attention and recurrence can be useful together. The study measured the trade offs explicitly: on MMLU the hybrid reached the same...

#ModelTraining #Multimodal #OpenSource #AI #AIPulse

A poem I wrote over the summer titled "Yellow Roses," accompanied by freehand pen sketches :)

#poem #poetry #poet #art #artist #artistsonbluesky #artistsofbluesky #pen #penart #traditionalart #writer #multimodal #plants #nature #romance

Training-Free Pseudo-Fusion for Composed Image Retrieval with Diffusion Models and Multimodal Lar...

Fan XU, Luis A. Leiva

Action editor: Xiao Luo

https://openreview.net/forum?id=6W3pFEQXZc

#multimodal #generative #retrieval

🤖 Cloudflare Unveils Multimodal Decision Model Clef-omni

Cloudflare's announcement expands its open weight decision model family by adding audio and video input alongside text and images, inside a unified sequence with video and audio...

#Multimodal #OpenSource #SpeechAudio #AI #AIPulse

Simple is Better than Complex: A Representation-centric Perspective for Prompting-based Vision--L...

Yujia Yin, Jinhong Ni, Renjie Wu, HONGJI LI, Tianxin Wei, Zhong Li, Yifan Chen

Action editor: Jaeho Lee

https://openreview.net/forum?id=yBVwYxHxUq

#attention #multimodal #fusion

🤖 Sharpa's D01 Robot Adds Touch to Its World Model

The description is a catalog of specifications and then a single sentence underneath: the robot's arm is built to carry its own weight, it can move at 10.5 metres per second, and it...

#Robotics #ComputerVision #Multimodal #AI #AIPulse

Didn't log #WeekWithoutDriving activities on the daily while I'm on vacation. I've been car & air passenger, rode bus & Montréal Metro, walked pedestrianized streets in Montréal. I get to see so much more at human speed! Public art, shops, people enjoying life. Zero parking hassle/cost. #multimodal

Discrete Diffusion in Large Language and Multimodal Models: A Survey

Runpeng Yu, Qi Li, Xinchao Wang

Action editor: Shuangfei Zhai

https://openreview.net/forum?id=0DsqnkP8Cp

#multimodal #decoding #dmllm

EmbeddingGemma 2：テキスト・画像・音声・動画を統合するオープンな軽量マルチモーダル埋め込みモデル

Google DeepMindが発表した「EmbeddingGemma 2」は、テキスト・画像・音声・動画を単一の空間にマッピングする7.4億パラメータのオープンな軽量マルチモーダル埋め込みモデルです。

#Multimodal #Embeddings #OpenSource #EdgeAI #RAG

🤖 Google's Compact EmbeddingGemma 2 Model Outperforms Larger Rivals

EmbeddingGemma 2 puts 740 million parameters into a model that converts text, images, video, audio and code into vectors, and claims it is the...

#RAGEmbeddings #Multimodal #BenchmarksEvaluation #AI #AIPulse

1/ #OpenCID : Ensuring total sovereignty for research data!
🔹 Local import & #multimodal management 🔹 #Python modularity (easy new formats) 🔹 100% Web-based & remote access 🔹 Auto-metadata extraction ( #FAIR / #OpenData) 🔹 Background imports for seamless workflow ⚙️

#DataScience #DataManagement