even if its not that interesting in that way im about to go bananas on multimodal
November 2, 2026 at 11:32 PM
even if its not that interesting in that way im about to go bananas on multimodal
Longfellow bridge multimodal sunset @universalhub.com
October 30, 2026 at 2:25 PM
Longfellow bridge multimodal sunset @universalhub.com
The Canadian Pediatric Imaging Platform (C-PIP) Reference Sample: Neuroimaging Protocol and Cross-site Validation for Harmonized Multimodal Pediatric Imaging | bioRxiv www.biorxiv.org/content/10.6...
November 2, 2026 at 1:53 AM
The Canadian Pediatric Imaging Platform (C-PIP) Reference Sample: Neuroimaging Protocol and Cross-site Validation for Harmonized Multimodal Pediatric Imaging | bioRxiv www.biorxiv.org/content/10.6...
An Editors' Pick via #OPG_BOEx: Multimodal 3D light-field and laser-speckle endoscopy https://bit.ly/48mdjDs #MinimallyInvasiveSurgery #OpticalImaging 💡🧪
November 2, 2026 at 6:05 PM
An Editors' Pick via #OPG_BOEx: Multimodal 3D light-field and laser-speckle endoscopy https://bit.ly/48mdjDs #MinimallyInvasiveSurgery #OpticalImaging 💡🧪
Beyond Multimodal Emotion Conflict: Temporal Context-Aware Hidden Emotion Detection with Multimodal Large Language Models
Yonghao Wang, Jiaxin Zhou
#arXiv #cs.AI
Yonghao Wang, Jiaxin Zhou
#arXiv #cs.AI
Beyond Multimodal Emotion Conflict: Temporal Context-Aware Hidden Emotion Detection with Multimodal Large Language Models
Multimodal emotion recognition commonly combines textual, acoustic, and visual signals to infer affective states. Recent work has begun to study conflicts among modalities, but a conflict does not necessarily indicate that a person is concealing an emotion. It may also result from mixed emotions, i…
arxiv.org
November 3, 2026 at 10:51 AM
Another exciting model for #openaccess multimodal publishing. Congratulations to @ox.ac.uk author @dacedzenovska.bsky.social and the whole team behind LIVING EMPTINESS. #collaboration #digitalpublishing
I'm beyond happy to announce the newest digital publication from @stanfordpress.bsky.social, Living Emptiness: Place, Power, and Meaning-Making from the Baltic to the Russian Far East. livingemptiness.org #supDigital
Living Emptiness: Place, Power, and Meaning-Making from the Baltic to the Russian Far East
What happens to places when people leave, services disappear, infrastructure decays, and value drains away? In <i>Living Emptiness: Place, Power, and Meaning-Making from the Baltic to the Russian Far ...
livingemptiness.org
November 3, 2026 at 4:04 PM
Another exciting model for #openaccess multimodal publishing. Congratulations to @ox.ac.uk author @dacedzenovska.bsky.social and the whole team behind LIVING EMPTINESS. #collaboration #digitalpublishing
Mistral has released Large 4, an open-weight multimodal model focused on coding, cybersecurity, agentic work, geospatial analysis, and 160+ languages alternativeto.net/news/2026/1...
November 2, 2026 at 10:04 PM
Mistral has released Large 4, an open-weight multimodal model focused on coding, cybersecurity, agentic work, geospatial analysis, and 160+ languages alternativeto.net/news/2026/1...
Multimodal AI can connect text, images, and audio in useful ways. The next step is designing interfaces that make uncertainty and source evidence easy to inspect. #AI #Multimodal #UX
November 3, 2026 at 9:17 AM
Multimodal AI can connect text, images, and audio in useful ways. The next step is designing interfaces that make uncertainty and source evidence easy to inspect. #AI #Multimodal #UX
CellWhisperer uses multimodal learning of transcriptomes and text to answer questions about single-cell RNA-sequencing data go.nature.com/3WPGJnW
rdcu.be/TamrmUPBYm6a
rdcu.be/TamrmUPBYm6a
Multimodal learning enables chat-based exploration of single-cell data - Nature Biotechnology
CellWhisperer uses multimodal learning of transcriptomes and text to answer questions about single-cell RNA-sequencing data.
go.nature.com
October 30, 2026 at 1:04 AM
CellWhisperer uses multimodal learning of transcriptomes and text to answer questions about single-cell RNA-sequencing data go.nature.com/3WPGJnW
rdcu.be/TamrmUPBYm6a
rdcu.be/TamrmUPBYm6a
There’s nothing as viscerally elucidating as trying to bike anywhere between the two parts of Alexandria. Any opponents to multimodal improvements should try it. If they survive, they’d be dyed in the wool transit reformers.
November 3, 2026 at 12:31 AM
There’s nothing as viscerally elucidating as trying to bike anywhere between the two parts of Alexandria. Any opponents to multimodal improvements should try it. If they survive, they’d be dyed in the wool transit reformers.
SenseNova-Loop-9B-Preview 👀 A looped multimodal model from SenseNova
huggingface.co/sensenova/Se...
- 9B
- Apache 2.0
- 56 layer executions via weight sharing
- Up to 256K context
- Hybrid attention + two stream mHC
huggingface.co/sensenova/Se...
- 9B
- Apache 2.0
- 56 layer executions via weight sharing
- Up to 256K context
- Hybrid attention + two stream mHC
November 2, 2026 at 1:32 PM
SenseNova-Loop-9B-Preview 👀 A looped multimodal model from SenseNova
huggingface.co/sensenova/Se...
- 9B
- Apache 2.0
- 56 layer executions via weight sharing
- Up to 256K context
- Hybrid attention + two stream mHC
huggingface.co/sensenova/Se...
- 9B
- Apache 2.0
- 56 layer executions via weight sharing
- Up to 256K context
- Hybrid attention + two stream mHC
PGGB improves multimodal learning by enhancing weaker modalities without reducing the power of stronger ones. Testing shows it effectively reduces the utilization gap, outperforming existing strategies in imbalanced contexts and remaining robust in balanced ones. https://arxiv.org/abs/2611.02651
Probe-Guided Gradient Balancing for Multimodal Learning
ArXiv link for Probe-Guided Gradient Balancing for Multimodal Learning
arxiv.org
November 3, 2026 at 9:00 PM
PGGB improves multimodal learning by enhancing weaker modalities without reducing the power of stronger ones. Testing shows it effectively reduces the utilization gap, outperforming existing strategies in imbalanced contexts and remaining robust in balanced ones. https://arxiv.org/abs/2611.02651
Our featured article: AI-based multimodal integration of genomics and electronic health records
go.nature.com/3WIWCzy
by Rasika Venkatesh & @marylynritchie.bsky.social
go.nature.com/3WIWCzy
by Rasika Venkatesh & @marylynritchie.bsky.social
AI-based multimodal integration of genomics and electronic health records - Nature Reviews Genetics
This Review covers advances in artificial intelligence-based methods for multimodal modelling of genomics and electronic health record data, and the integration of these data into electronic-health-re...
go.nature.com
November 4, 2026 at 9:55 AM
Our featured article: AI-based multimodal integration of genomics and electronic health records
go.nature.com/3WIWCzy
by Rasika Venkatesh & @marylynritchie.bsky.social
go.nature.com/3WIWCzy
by Rasika Venkatesh & @marylynritchie.bsky.social
Sensorium: An Open-Source Multimodal Data Collection and Curation Library for Embodied Surgical AI
Lorenzo Mazza, Ariel Rodriguez, Stefanie Speidel
#arXiv #cs.AI #cs.RO
Lorenzo Mazza, Ariel Rodriguez, Stefanie Speidel
#arXiv #cs.AI #cs.RO
Sensorium: An Open-Source Multimodal Data Collection and Curation Library for Embodied Surgical AI
Embodied surgical AI requires reliable multimodal data, yet surgical data collection remains costly, difficult to repeat, and strongly dependent on the hardware setup. We present Sensorium, an open-source Python library for recording, processing, annotating, and curating multimodal surgical data fr…
arxiv.org
November 3, 2026 at 12:08 AM
Multimodal, end-to-end encrypted, stateless, portable, single-binary, multi-threaded piss bottles under the desk
November 3, 2026 at 7:47 PM
Multimodal, end-to-end encrypted, stateless, portable, single-binary, multi-threaded piss bottles under the desk
SenseTime's deep processing trick
SenseTime released SenseNova Loop-9B-Preview, an open-source 9B multimodal model that doubles its processing depth by reusing a shared block of layers.
SenseTime released SenseNova Loop-9B-Preview, an open-source 9B multimodal model that doubles its processing depth by reusing a shared block of layers.
November 2, 2026 at 8:01 AM
SenseTime's deep processing trick
SenseTime released SenseNova Loop-9B-Preview, an open-source 9B multimodal model that doubles its processing depth by reusing a shared block of layers.
SenseTime released SenseNova Loop-9B-Preview, an open-source 9B multimodal model that doubles its processing depth by reusing a shared block of layers.
Multimodal characterization of putative pan-inhibitory viral constructs in macaque monkeys https://www.biorxiv.org/content/10.64898/2026.10.27.761178v1
November 3, 2026 at 2:15 PM
Multimodal characterization of putative pan-inhibitory viral constructs in macaque monkeys https://www.biorxiv.org/content/10.64898/2026.10.27.761178v1
MAGAR: Multimodal Agents for GUI-Based 3D Mechanical Assembly and Minimal Rework
Jingxuan Wei et al.
#arXiv #cs.AI
Jingxuan Wei et al.
#arXiv #cs.AI
MAGAR: Multimodal Agents for GUI-Based 3D Mechanical Assembly and Minimal Rework
A GUI command can be correct at the interface level yet still produce an incorrect assembly. An agent must identify the relevant components, respect their structural relationships, and determine which parts should move during recovery. We introduce MAGAR-Bench, a benchmark of 4,000 multimodal sampl…
arxiv.org
November 3, 2026 at 8:38 PM
Mistral Large 4 drops weights 🦾
Mistral AI has released the weights for its 1-trillion-parameter multimodal model, Mistral Large 4, on October 29, 2026, reportedly leading open-weight coding benchmarks.
Mistral AI has released the weights for its 1-trillion-parameter multimodal model, Mistral Large 4, on October 29, 2026, reportedly leading open-weight coding benchmarks.
November 3, 2026 at 11:04 AM
Mistral Large 4 drops weights 🦾
Mistral AI has released the weights for its 1-trillion-parameter multimodal model, Mistral Large 4, on October 29, 2026, reportedly leading open-weight coding benchmarks.
Mistral AI has released the weights for its 1-trillion-parameter multimodal model, Mistral Large 4, on October 29, 2026, reportedly leading open-weight coding benchmarks.
Crouton Alpha drops on OpenRouter 🔎
A new anonymous model, Crouton Alpha, appeared on OpenRouter yesterday, offering a 1M token context window and multimodal input for free.
Another "stealth" model provides free load testing for an undisclosed major lab.
A new anonymous model, Crouton Alpha, appeared on OpenRouter yesterday, offering a 1M token context window and multimodal input for free.
Another "stealth" model provides free load testing for an undisclosed major lab.
November 4, 2026 at 10:53 AM
Crouton Alpha drops on OpenRouter 🔎
A new anonymous model, Crouton Alpha, appeared on OpenRouter yesterday, offering a 1M token context window and multimodal input for free.
Another "stealth" model provides free load testing for an undisclosed major lab.
A new anonymous model, Crouton Alpha, appeared on OpenRouter yesterday, offering a 1M token context window and multimodal input for free.
Another "stealth" model provides free load testing for an undisclosed major lab.
PhenoMe #standards facilitate the integration of additional biological readouts into the image datasets, allowing a joint (multimodal😉) exploration and analysis
November 4, 2026 at 5:06 PM
PhenoMe #standards facilitate the integration of additional biological readouts into the image datasets, allowing a joint (multimodal😉) exploration and analysis
