ECMWF models predict a massive low latitude Western Distrubance... to affect india during first week of December. Almost 10 days to go so models may change however remain watchful for this remote possibility.. Video Courtesy Windyshow more

Massive Western Distrubance likely to impact north India anytime... during 18-23 January. Remain Watchful. Video Courtesy Windyshow more

A gigantic Western Disturbance is likely to affect parts... of north India during 27-29 January. Significant rainfall or hailstorms are expected over Uttar Pradesh, Uttarakhand, Haryana, Delhi NCR, north MP, and parts of Himachal Pradesh and Punjab. Parts of Jammu Kashmir, Himachal may also experience rainfall and snowfall. Remain watchful. Video Courtesy Windyshow more

BREAKING - Almost 900 migrants cross the Channel illegally... so far this weekend, taking the total number of arrivals this year above 33,000. After a week of bad weather, conditions are expected to remain fairly good for the next few days, with more crossings expected.show more

Predicting the next word "only" is sufficient for language... models to learn a large body of knowledge that enables then to code, answer questions, understand many topics, chat, and so on. This is clear to many researchers now, and there are nice tutorials on why this works by Ilya Sutskever resorting to compression ( ) and by Geoffrey Hinton ( ). However, the emergence of types of understanding is not unique to language models. In by Misha Denil and Brandon Amos the authors trained models to predict the next few time stems of over a hundred robot hand sensors (Touch, Gyro, Accelerometer, Joint Info, Actuator Info, etc.). They ten found out that they could regress the shape of the thing the hand was touching from the activations of the neural networks using probes. That is, the model developed an internal representation of shapes even though it was simply used to predict "only" the next few senses. Awareness follows from simple predictions and interaction with the world.show more

Brisbane (Australia) is expected to be directly hit by... a tropical cyclone #Alfred 🙏🧐 The storm will slow down significantly before making landfall, thus resulting in extremely heavy rainfall along the coast. ICON models predict over 500 mm of precipitation in some areas by the end of the week. Flooding is expected. The video shows the progress of the storm over the next 3 days:show more

We’re not prepared for how quickly production pipelines are... going to change with AI. Some of the latest video models have immediate implications for Hollywood - endless character swaps at a negligible cost. (this is from ederxavier3d on IG using Kling Motion Control)show more

GFS has gone crazy. It is forecasting an LPS... entering through Odisha and moving toward Maharashtra around May 25-27. It's a very rare chance, but it indicates large-scale flow patterns are stabilizing in a way that is favorable for an LPS to enter deep into India's interior by end of this month. Video Courtesy Windyshow more

Exactly 1 year ago I started March with my... last $100 I lost almost all my money on dumb decisions / overtrading. Feels crazy to look back to these times, so much that can change in 1 year don’t take it for granted. This is just the first 4 days of March tho, way more to come.show more

FDA HAS RELEASED A NEW GUIDELINE: IF YOU HAVE... ACCIDENTALLY DIGESTED A TOXIC SUBSTANCE, IMMEDIATELY WATCH THIS VIDEO TO INDUCE VOMITING. ⚠️WARNING⚠️ VOMITING CAN LAST FOR UP TO 10 DAYS. SIDE EFFECTS MAY INCLUDE NAUSEA, NIGHTMARES, AND A PERMANENT LOSS OF APPETITE.show more

UPDATE: A LOT of tropical moisture on the way... across Florida the next 5-10 days. For example, those moisture levels will be 50%-75% above average, which will aid in developing numerous rounds of storms across the state. Values shown are PWATS or precipitable water levels. Around 1.6" is typical this time of the year for Florida, which you'll notice we'll be above the next 7-10 days. Some online chatter about tropical development. Not impossible, but current odds are low. Physics-based models show the potential, AI models largely don't. Rainmaker regardless if it tries to form into a weak tropical/subtropical system or not. We need the rain!show more

Rerender A Video: Zero-Shot Text-Guided Video-to-Video Translation paper page:... Large text-to-image diffusion models have exhibited impressive proficiency in generating high-quality images. However, when applying these models to video domain, ensuring temporal consistency across video frames remains a formidable challenge. This paper proposes a novel zero-shot text-guided video-to-video translation framework to adapt image models to videos. The framework includes two parts: key frame translation and full video translation. The first part uses an adapted diffusion model to generate key frames, with hierarchical cross-frame constraints applied to enforce coherence in shapes, textures and colors. The second part propagates the key frames to other frames with temporal-aware patch matching and frame blending. Our framework achieves global style and local texture temporal consistency at a low cost (without re-training or optimization). The adaptation is compatible with existing image diffusion techniques, allowing our framework to take advantage of them, such as customizing a specific subject with LoRA, and introducing extra spatial guidance with ControlNet. Extensive experimental results demonstrate the effectiveness of our proposed framework over existing methods in rendering high-quality and temporally-coherent videos.show more

I explored a further possibility with local models: Qwen3.6... 35B A3B + NVIDIA LocateAnything-3B as a local Computer Use agent (proof of concept). In the demo, I asked it to switch my Mac to light mode. It did. Then back to dark. Did that too — finding the right toggle in System Settings, clicking it, and verifying the change itself. It's fully screenshot-based, so no Accessibility API needed. If it's on screen, the agent can see it and act on it. This runs entirely on your own hardware — private, local, built from two small open models.show more

It's been so hard to get anywork done. I... have gotten sick twice in the past 2 weeks. Still recovering but it inspired me to create this cute lil care kit for vtuber models. Still a WIP. But hopefully will be available before the end of the week. ♡ #vtuberartist #vtuberassets #live2dwipshow more

i’ll always be so deeply interested in the early... days of this fandom. i would really love to interview a select group of fans that truly experienced what it was like during the early stages and what it felt like to begin to see fan videos and content for the first timeshow more

This clip from the Visma Classics recap highlights Wout's... issue. Due to his crashes, and as he gets older, his tolerance for risk is dropping If it continues to be a problem, he may not be able to support Vingegaard during the first week of the Tour, as it will be super hecticshow more

✨ Grok's new Imagine video model also comes with... an Edit model We know edit models for images, you submit an image, write a prompt what to change, but this is the first time I've seen a proper edit model for video And it kinda works, not great yet though but it does something Here I had to remove the old name "Nomad List" in the video for my site First it said "Go nomad -> Nomad List", so I prompted it "remove the text Nomad List. do not change anything else", it didn't remove it but it replaced it with just "Go nomad" again, okay good enough Useful because otherwise I'd have to scour my backups for the original video in Final Cut Pro, and this is faster One thing you see is it changes the pattern on the door also, but that's okay for now if I fade it inshow more

A major question in multimodal modeling is how to... leverage strong pre-trained models, such as industry-scale LLMs and VLMs, during training to avoid starting from scratch. This is important as the demand for rich and domain-specific multimodal models continues to increase, while training them from scratch is obviously impractical due to limited data and compute. The so-called “any-to-any” multimodal models can model a large diverse dictionary of modalities. That’s good. Their downside is that their architectures are often not decoder-only, which has limited their performance in practice and prevents them from leveraging strong pre-trained decoder-only models as priors. We are releasing MODUS (ICML Conference '26) a decoder-only any-to-any multimodal model to address some of these questions. A single transformer decoder predicts any modality from any others with no modality-specific heads, losses, or task pipelines. We show efficient adaptation of established models, e.g., BAGEL, to rich any-to-any multimodal modeling. We are releasing 14B to 77B parameter models. All materials are open-source. The download links and demos here 🧵show more

I take it back, this video may take a... little while longer to be complete. I'm not too pleased with some of the framing of some shots and theres so much that still need to be edited in order for me to be proud of it. For now, here's a rough first look at: Graveyard Shiftshow more

During the crustal displacement described in the censored book... The Adam and Eve Story: A History of Cataclysms by Chan Thomas, who worked for a company contracted by the Air Force in 1963, being aboard a doomsday plane could actually be one of the safest places, assuming the low magnetic field would not affect the aircraft. The spilling of the oceans would cause tsunamis up to 2 miles high, reaching even high-altitude mountain doomsday bunkers. This is also one of the reasons for developing reusable rockets: so that a number of elites could remain in space for a period of time until everything settles down again.show more