Nebsh's Compact LoRA Brings Seedream 5.5 Lighting to Open MiniMax H3
A community-trained LoRA adapter pushes MiniMax H3 Image toward Seedream 5.5's cinematic photographic look, with a 1,134-image dataset and 6,000 training steps.
20 stories
A community-trained LoRA adapter pushes MiniMax H3 Image toward Seedream 5.5's cinematic photographic look, with a 1,134-image dataset and 6,000 training steps.
Kyutai's AURA distills a bidirectional video diffusion teacher into a causal student that streams lip-synced talking heads live on one H100.
Google's Genie 4 world model is now open to paid Gemini API users. It streams 720p worlds with native audio that you can steer live from code, at about $0.10 per second.
Goodfire mapped every neuron in Qwen3.6, arguing that individual units, not learned features, are the right basis for interpretability.
Google's Antigravity desktop app can now open a remote Linux machine over SSH, so agents, terminals and Git worktrees run on the server and keep going after you close the window.
Parcha Labs open-sourced AgentRun, a layer that sits around Claude Code, Codex, Pi or OpenCode and adds learning, crash recovery and swappable computers, so repeated agent work gets cheaper.
Google's fast Gemini 4 Hydrogen model now comes with an open-source Antigravity Robotics plugin. It handles robot setup, MuJoCo simulation and camera-guided control from plain-language prompts.
A co-designed FPGA neural decoder called Rapids keeps pace with superconducting qubits while cutting logical errors 240 times below prior real-time decoders.
Microsoft's new Canvas workspace runs on MAI Image 2.7 Flash. You can sketch, select and edit parts of an image in one place, at about a cent per image through the API.
Nine frontier decision models were given three in-game years to grow a RollerCoaster Tycoon 2 park to 1,000 guests. Six failed.
Perplexity's Brain now mines your past sessions to auto-write reusable agent skills, cutting tool-call failures by 11% in production tests.
The community ComfyUI port of ByteDance's one-step SeedVR2 restorer passes 800K monthly downloads. It runs 7B video upscaling on 8GB GPUs and is free.
Krea opens its creative agent beta to more testers after 50,000 signups, handing out $100 credits to brief a crew of specialist bots.
Pruna AI turned MiniMax's 143 GB H3 video model into an open image generator and editor that renders 8 MP images in 8 steps and fits in 21 GB of VRAM.
Meta's new lldb-server plugin treats CPUs and GPUs as separate debugger processes in one session, with most of the framework now upstream in LLVM.
Linger, a new open-source personal agent, runs on consumer PCs. It handles memory, agent teams and email triage locally using a MoE engine that fits a 125B model onto a 12 GB GPU.
A NeurIPS paper looks inside looped language models and finds that each pass through the shared block repeats the same processing stages a normal Transformer runs once. Whether those stages hold steady depends on a few design choices.
Fast-JEPA rearranges how latent world models search over action plans, cutting planning time to 10.88 ms while hitting 96.50% average success.
Lastbook packages a local Gemma model, offline Wikipedia, street maps and survival videos into one private desktop app for Mac, Windows and Linux.
Monash and German researchers prune the Hydra and Quant time series classifiers, cutting inference energy by up to 70% with only a small loss in accuracy, and show that hardware choice can change which model looks best.