Loop | Cleaned | A Steamy Night With Yae... Miko and Raiden [Highly requested, so I have removed the fog and blur. 🫡] Models by DaB #nsfwtwt #nsfw #RaidenShogun #RaidenEi #yaemiko #yaemikonsfw #genshintwt #GenshinImpact #blender #animation3dshow more

Saucifer 🔞
70,979 görüntüleme • 1 yıl önce
🚁 Adding an Apache attack helicopter today I think... I will rewrite the entire flight model today again to be ambiguous of what kind of vehicle it is Planes behave very different than helicopters and different than vehicles like tanks or pedestrians Because people here also requested "Guy with bazooka" so that will be next I have no idea if everyone else is just lying about using AI for their plane models but mine look so clunky (but I like the look), but I now tried with Claude 3.5 and 3.7 and Grok 3 and they all just generate clunky models, not beautiful like you see in other AI vibe coded games, so either they bs'ing or I'm retarded I also have to move sub parts of each model like for 30 minutes until everything fits, AI has a really hard time with 3d models But okay the look is kinda cuteshow more

@levelsio
485,431 görüntüleme • 1 yıl önce
I lowkey built a $11M/month app idea with just... 4 prompts Didn’t expect this many people to actually care So let’s talk about the updates I’ve made over the last hour The first version worked But it felt static "Didn't have a soul" And static apps don’t retain users So I made a few key changes: → tightened the core loop → removed friction from the first action → had Rork generate me better images So in 6 more prompts, 10 total That made the app feel 10x better. Building is solved. Iteration speed is the advantage now. Question remains, Should I ship?show more

Jah Jiren
66,438 görüntüleme • 7 ay önce
SUNNY HILL FESTIVAL NIGHT ONE 🇽🇰❤️ This festival is... something my family and I created with so much love for our city PRISHTINA and to see it come to life like this is beyond anything I could’ve dreamed of. Sharing this night with you all, in the city that shaped me, surrounded by so much energy, joy, and pride… it’s hard to put into words what it means. Performing for you, and sharing the stage with my dad Dukagjin Lipa was a moment I’ll never forget. I feel so incredibly lucky and grateful. And we still have two more nights to celebrate together!! JU DUA me krejt zemrën ❤️❤️❤️❤️❤️show more

DUA LIPA
194,520 görüntüleme • 1 yıl önce
I've used ghee a fair bit in my animal-based... life but I was never able to tolerate the taste of it straight from the jar. This Great Ghee product is different. It tastes more like butter than any ghee I have ever used plus it's lab tested for glyphosate, grass fed and made from A2 cows. Ghee is clarified butter with the water and milk solids removed, leaving it almost pure fat. It is a low omega 6 concentration of saturated and monounsaturated fats and has little lactose and casein, which is why many people tolerate it better than butter and it's why it has a very high smoke point. Highly recommend this ghee and it's made by a small business in Virginia.show more

Cary Kelly
25,775 görüntüleme • 5 ay önce
New open-source agent harness just landed! I got early... access to TrueForge by TrueFoundry and have been running it locally for the past few days. The harness layer deserves as much attention as the model, and open source matters here because you can inspect the loop, run it on your own infrastructure, and swap to the latest or cheaper models. TrueForge handles the runtime work that makes an agent reliable. It drives the tool-calling loop, manages context, coordinates subagents, and executes code in a sandbox, with any model you choose. Every tool call re-sends the growing context to the model, so in practice the harness controls most of what an agent costs to run. A few things stood out from my testing and their published benchmarks. Vendor-Neutral by design. It runs OpenAI, Anthropic, and Google models alongside open-weight models like Kimi, GLM, and DeepSeek. Model routing is a setting, and you can send each task to the model that fits it. On a 14-task enterprise agent benchmark, it matched the accuracy of Claude Managed Agents running the same Opus 4.8 model at roughly 30% lower cost per run (3.8M tokens vs 10M for the same answers). Routing the same tasks to GLM-5.2 held accuracy and brought cost down by about 75%, around $3 per run instead of $12. Fully self-hosted and Open Source (MIT License). I had it running locally with one command, with sandboxed code execution working out of the box. It's time to own your agent harness. Thanks to TrueFoundry for partnering on this post.show more

elvis
11,303 görüntüleme • 1 gün önce
Don't train the model, evolve the harness. I read... a brilliant blog post from Hugging Face where they took a frozen open model scoring 0% on a hard legal agent benchmark, left its weights alone, and let an automated loop rewrite only the code around it. That code layer is the harness, the runtime wrapper that feeds the model context, runs its tool calls, and decides when a run ends. By the time the loop finished, the system had essentially matched Sonnet 4.6 on the benchmark's headline metric, at roughly 7x lower cost per task. Zero weights changed. The gain existed because of where the model was failing. The judge only grades files saved in the right place under the exact requested filename, and the model kept doing the legal analysis correctly, then saving it under the wrong name, dropping it in a scratch folder, or never writing it at all. So the 0% was never measuring legal reasoning. It was measuring the harness. Hand-tuning that layer is slow and model-specific, so they automated it. A Claude proposer adds exactly one mechanism per iteration, and an outer loop keeps it only if it clearly beats the current best, so accepted mechanisms compound. What the loop discovered says a lot about where agents actually fail. → The biggest single gain was file handling, not intelligence. An automatic step that lands the deliverable exactly where the judge expects it beat every prompt change, with zero extra model tokens. → Code fixes transferred across models, prompt playbooks did not. The same harness lifted a smaller model from the same family by 14 points, but the tuned prompts hurt a different model family on tasks it could already finish. → The harness mattered more than anything else. Same model, same judge, same tasks, and five different harnesses scored anywhere between 3.5% and 80.1%. The gains do eventually flatten, and the remaining misses look like real capability gaps. At some point the wrapper runs out of tricks and the model has to carry the work. But the lesson holds. A benchmark score measures the model and its harness together, and until the harness is fixed, it's impossible to know which one failed. I highly recommend reading this: I also wrote a deep dive on agent harness engineering a while back, covering the orchestration loop, tools, memory, context management, and everything that turns a stateless LLM into a capable agent. The article is quoted below.show more

Akshay 🚀
244,567 görüntüleme • 1 ay önce
aespa danced supernova at the wedding bc it was... requested by the newlyweds 😂 karina: they said crying at a wedding makes the atmosphere awkward, but we (aespa) have so much affection for unnie (bride is their hair stylist) and oppa (groom is their creative director), i guess we just got overwhelmed. we prepared a gift that the newlyweds wanted to hear, if you happen to know (supernova), we would love if you could sing along with usshow more

Sunshine
181,915 görüntüleme • 3 ay önce
We are stoked to formally announce The Chefs. A... fully dynamic PFP collection with art created by Rugdrigo (Artist 2.0). Inspired by what Sanj did with Fomojis 2.0s, we decided to take it a step further and have all of the traits evolve from day to night. So the day time versions of your ordinal will be chefs. The night time versions will evolve into something else entirely. With our 1/1s, we wanted to do something a bit different. I believe 1/1s should really stand out in a collection and so few projects actually succeed in doing that. I reached out to some of my favorite artists and projects in the space to curate an incredibly diverse set of 1/1s. We do the cooking! Mint Date : 11/15 on Magic Eden on Bitcoin 🟧 at 10 AM EST Price : .0009BTC Collection Size : 2392show more

Jalapeno Pete
12,657 görüntüleme • 1 yıl önce
There's something about design that truly makes me happy.... Before I was diagnosed with a complete SCI, I loved building thing with my hands and trying to make them unique and beautiful. The accident didn't just take my mobility, it robbed me of my creative drive. I didn't I would ever get that back that was up until Neuralink came into the picture. I was very serious about applying but just couldn't bring myself to do it. Then Noland Arbaugh comes into the picture and I see his video and I saw how passionate he was toward the whole ide and that was like a switch. I want to thank Noland because if I hadn't seen that video I wouldn't be typing this message with my thoughts. So the day I gained control of the device I had the guys install Autodesk Fusion and the first night I had my first part designed by BCI. I have been designing things non-stop since. Now lets talk tools and the first one is the ArcDroid it is a CNC robot that holds a plasma torch and can cut profiles in sheet metal. I have made numerous signs, letters and parts with this and i love it. I have also had the chance to help the local high school install and setup their CNC plasma table. I love being able to share my knowledge with others and show new technology to kids that want to learn. The next one everybody is more familiar with is the Bambu Lab X1C. It is my first 3D printer and I had no idea how dang handy these things were! I can go from thought, Neuralink, Fusion, X1C and poof! You have your part all in a matter of like a hour. I make everything from parts for my wheelchair or RC ✈ to toys for my nephew. These devices have completely changed my life. I now have my creative outlet back and I'm not done looking for new things to try. I'm thinking a laser next!!!😎 Thank you Elon Musk for the opportunity to showcase my skills, I'm extremely grateful🙏show more

Alex Conley
43,158 görüntüleme • 8 ay önce
Ignition: Powered by Seamless Motion🔥 A cinematic journey from... spark to speed. Smooth transitions and flawless motion bring the engine’s awakening to life, showcasing the incredible cinematic flow made possible with Kling. GPT Image 2.0 x Kling 3.0 on Yapper Prompt: Ultra-cinematic 10-second photorealistic sequence of a futuristic high-performance sports car being powered on and driving away. Smooth motion, dramatic storytelling, 24fps, HDR, shallow depth of field, film grain, anamorphic lens flare, volumetric fog, high contrast lighting, 4K. Mysterious driver wearing a modern black racing jacket and black gloves, face mostly hidden in shadow and rim lighting, confident calm movements, realistic human proportions and textures. Environment: dark industrial underground garage at night with concrete walls, metal pipes, subtle neon accent lights, wet reflective floor, volumetric fog and cinematic light rays. Garage exit leads to a modern city street at night with distant streetlights and soft bokeh. START FRAME: wide cinematic shot of sleek futuristic sports car parked in the foggy garage, driver standing beside the car holding a glowing smart key, headlights OFF, neon reflections on glossy paint, wet reflective floor, dramatic composition. SEQUENCE ACTION (continuous cinematic flow): The driver presses the glowing ignition button on the smart key. Camera slowly dollies toward the car as the vehicle awakens. Headlights power on gradually, glowing through the fog, subtle vibration ripples across the car body, reflections shimmer across the paint. Seamless cinematic cut into extreme macro view inside the engine. Highly detailed mechanical components in slow motion pistons firing, turbines spinning, electric energy flowing through cables like glowing currents, sparks igniting, heat waves and light pulses building power. Camera smoothly pushes and orbits through engine parts as the energy intensifies, orange and blue glow illuminating metal textures. Cinematic cut back to low front angle of the car. Engine revs aggressively, headlights flare bright, dust and smoke swirl around the tires. The car launches forward smoothly and powerfully toward the exit. Camera tracks low as the car speeds past and exits the garage into the night street, motion blur and light streaks trailing behind. END FRAME: rear three-quarter view of the sports car speeding away into the night city, glowing red taillights leaving light streaks, drifting smoke fading behind, cinematic fade to black. Style keywords: cinematic, photorealistic, smooth tracking shot, macro engine detail, futuristic technology, dramatic lighting, epic atmosphere, high dynamic range.show more

Jayden™
15,409 görüntüleme • 3 ay önce
Throughout my journey in developing multimodal models, I’ve always... wanted a framework that lets me plug & play modality encoders/decoders on top of an auto-regressive LLM. I want to prototype fast, try new architectures, and have my demo files scale effortlessly — with full support for parallelism and optimization. Not just to hack⚙️, but also to scale🚀. So finally we built it for ourselves. LMMs-Engine: a lean, efficient framework built to train unified multimodal model at scale. From Qwen LLM, VLM, LLaVA-OV, and WanVideo, to unified models like Qwen-Omni and BAGEL — plus Linear-Attn GDN and research prototypes like RAE and SiT - all under one modular system that seamlessly integrates diverse datasets and optimization strategies. Powered by FSDP2 multi-dim parallelism, Ulysses sequence parallel, Flash-Attention, Liger Kernels, and Native Sparse Attention (also with bonus support for the Muon optimizer for all models).show more

Brian Li
54,840 görüntüleme • 9 ay önce
I was visiting Goa with my wife a couple... of years ago. We were in this cosy little resort in South Goa right on Agonda beach, run by a middle aged Goanese couple. We had stayed there for three days and on the 4th day we drove to North Goa, walked into a bar and got drunk and so had to check in at a different hotel for the night. At 11:30 ish, the lady from the resort called us to check if we were okay and that she had waited up to see if we would return. I thanked her and assured her that everything is okay. I don’t think I have or will experience this anywhere else, and I am a fairly well traveled person. India, with all its follies, can sometimes give you the most heartwarming experiences. This was the view from our room at the resort.show more

Mayukh
694,503 görüntüleme • 7 ay önce
Excuse me for breaking the wall and sharing personal🤟... I just turned 40, which brought about a wave of reflection, and I'm incredibly lucky. I make my morning commute like most, grind at a desk until 5, then I flip roles to becoming a podcaster, a writer, a coach, a husband, and a dad. Life is moving at warp speed. I'm not rested, but I'm not tired. I'm not polished, but I'm not giving up. I'm not content, but I AM happy. The support and friendships I’ve found in this space have genuinely made my life better. This road, where passion meets persistence, has been paved by so many of you, and I’m incredibly grateful for that. I hope I’ve come across as someone reliable. Someone who isn’t afraid to put in the work to keep improving. Someone who cares, connects with people, and takes this seriously. I’ve always wanted to blur the lines between the person giving out golf picks on YouTube and the guy who simply wants the best for his family. At the end of the day, both are me. We took my son to his first concert. A random person took this video of us, found my wife, and sent this. As life enters the second half, this is how I want to be remembered. Unbothered by embarrassment, just going for it with the people I love, and on the forever pursuit of happiness 🏁 People told me slow my role...show more

Joe I
12,463 görüntüleme • 2 ay önce
One thing we are watching now is potential ERC... (eyewall replacement cycle) beginning. Pressure / wind may have temporarily peaked. Some of the hurricane models have suggested multiple eyewall replacement cycles through Tuesday. For tonight, it is possible this eyewall replacement is quick then we see another dip even stronger overnight with lower pressure and even higher winds. These eyewall replacement cycles will have impacts on how the storm comes into FL Wednesday night. Eventually these should grow the eye and core with time. If the cycles are finished before it hits the wind shear, that could make it a little more resilient to the wind shear. A smaller eye and core probably gets bothered more by the dry air intrusion/wind shear that should be meeting up with #Milton late Wednesday off the Tampa coast. So, these cycles are important in the eventual core structure as it nears the coast late WED.show more

Noah Bergren
111,627 görüntüleme • 1 yıl önce
I see only Akhil's story , no one is... ready to know about Mega's story. Why would a woman with fame, money will choose a bus driver 🤔instead of her husband ? Something must have gone wrong? .. so stop throwing abusive comments on a woman because it is easy . Thiruvananthapuram couple vloggers story - Akhil and Mega were a highly popular vlogger and influencer couple who had lakhs of fans on Kerala social media platforms. This young couple, who lived on the same street, fell in love and got married. They shared moments from their happy life on Instagram and YouTube and were regarded as the dream couple by their fans. In the past few days, a complaint had been filed at the police station stating that Akhil’s wife Mega had not been seen. Even after the intense investigation conducted by the police and examination of the CCTV footage in the area, shocking information came to light. It has been confirmed that Mega took all the items from her house, including the money box and valuable jewellery, and left the house at midnight with a local bus driver.show more

NoturEx
143,459 görüntüleme • 9 gün önce
UPDATE: June 2026 coverage by The Telegraph by Camilla... Turner I believe this is: *Part 1: *Part 2: will be in the Sunday Telegraph paper *Part 3: a longer video piece online next week. *************** I will be releasing the FULL audio meeting I had with Andy Burnham and IOPC Independent Office for Police Conduct (IOPC) Regional Director Catherine Bates. The public needs to see the level of the cover up and corruption involved in protecting predatory Greater Manchester Police officers and how I have been treated. This meeting exposes who is telling the truth...ME and who is lying...THEM! I was drugged and sexually assaulted in police custody in Feb 2021. I requested the footage and 4.5 hours worth of footage was concealed including the 'missing' EVIDENTIAL FOOTAGE. Currently the request for digital forensics which would prove what actually happened to the evidential footage has been withheld for 23 months by Greater Manchester Police Andy Burnham Independent Office for Police Conduct (IOPC) ICO - Information Commissioner's Office Why should I protect the establishment bodies when they have failed to protect me and the wider Greater Manchester public? 🩷THANK YOU to everyone who has followed my journey, believed me, supported me and helped me raise awareness despite the smear attempts🩷 I have data that is not currently in the public domain simply because its not the right time nor are the current police complaints bodies the right people. So, watch this space as there is more to come...😉show more

(was) Seeking Justice (now) Raising Awareness
39,599 görüntüleme • 2 ay önce
As always everyone is blind staring at the progress... of LLMs for coding and chat But meanwhile the new SOTA video model Seedance 2.5 has been slowly rolling out and it's really quite exceptional It's made by ByteDance (TikTok) who of course have lots of training data With just a few reference pics, it can get quite close to how you look IRL and you can do quite professional video shots with just a prompt I'd say it's the first video model that's now at the level of image models with the level of character likeness, cracking that in image models also took about 3 years (2022-2025) Generating 15 seconds takes about 4 minutes I put it live now on Photo AI, you can use it under [ Make video ] from the sidebar with just a prompt and your model selected! So you don't need to take an AI photo first and then turn that into a video! Saves lots of time :D It's more expensive than but I kept the credits the same (30 for 1 video) It also works inside the new video editor and you can make changes in your video with [ Magic edit ] in both the main app and the video editor Also a message for my server guy Daniel Lockyer (it can do voice too and you can even submit a voice sample of yourself, but I didn't here)show more

@levelsio
704,902 görüntüleme • 12 gün önce
Protesters at Saint Annes police station last night. I... have seen in the Echo today that politicians have come out condemning the next protest organised by the same person on County Road next week. Country Road is one of many areas in Liverpool that has undergone massive demographic and cultural change in a short space of time. Many of the locals are upset that this has been done to their home without consulting with any of them. It’s not that they were ignored. They were never part of the decision-making process to begin with. So they are taking to the streets to make themselves heard. And now they are being silenced by the very people that did this to them.show more

Around Liverpool
11,999 görüntüleme • 2 ay önce