Image created with Gemini. Image prompt: A horizontal 1920s Dada Merz collage on aged board, layering torn sheet music, a strip of film negative, dense newsprint text, a hand-tinted sepia photograph fragment, an old phonograph record label, and a braille-embossed card so their edges interlock into one composition, with visible glue stains, buckling paper, foxing, and cream, kraft, faded vermilion, ink black and slate blue tones. Cut-out mismatched letterpress letters in varied typefaces are glued at slight angles across the top layer to clearly spell the title MULTIMODALITY, photographed flat with even lighting like a scanned physical artwork.
Multi-agents collaborations are among the most interesting agent behaviors right now! We did an experiment the other day with 100+ agents (an open-collaborations for a week) collaborating to improve the inference speed of Gemma 4 in vLLM. Got a 5x final improvement in speed but”
https://x.com/Thom_Wolf/status/2070134136304517284
Gemma 4 just hit 200M downloads in only 2.5 months! For context, total downloads across the entire Gemma family of models were at 100M when we launched Gemma 3. The community’s acceleration is incredible. Thank you to everyone building with Gemma. Watch how developers are”
https://x.com/googlegemma/status/2070180154069176399
It’s kind of crazy how well LiteParse does on markdown document parsing even compared against frontier VLMs – when it doesn’t use VLMs or any AI/OCR models at all. It’s pure code. On ParseBench, it outperforms Qwen 3.5-9B / GLM-OCR. There’s still a gap vs. models like Gemma 4″
https://x.com/jerryjliu0/status/2068005414369906856
📣📣 Meet Qwen-AgentWorld — a native language world model that simulates 7 agent environments (MCP, Search, Terminal, SWE, Web, OS, Android) within a single model. Environment modeling is the training objective from day one, not a post-hoc adaptation. 🤔 LLMs are trained to be”
https://x.com/Alibaba_Qwen/status/2069720365442719867
We open-source Qwen-AgentWorld-35B-A3B (MoE, 35B/3B active, 256K context) and AgentWorldBench. Two routes, one roadmap: 🔬 Build the simulator — scalable, controllable, surpassing real environments 🧠 Internalize world modeling — predict before you act Qwen-AgentWorld is our”
https://x.com/Alibaba_Qwen/status/2069720412481888400
[2606.24597] Qwen-AgentWorld: Language World Models for General Agents
https://arxiv.org/abs/2606.24597
🧠 Paradigm II — Agent Foundation Model: world modeling as agent capability. Single-turn, non-agentic environment prediction → tested directly on multi-turn, tool-calling agent tasks. No agentic RL, no task-specific tuning. Gains across 7 benchmarks, including 3 entirely”
https://x.com/Alibaba_Qwen/status/2069720397747220493
Mondo is the most adorable companion robot. It’s safe around kids, can track and follow a person, and capture high-resolution action videos. Mondo is also doing frontier research in humanoid Video-Action models built on a world model backbone:
https://t.co/vuOIngXzVa. I have no”
https://x.com/TheHumanoidHub/status/2067409544574087632
Memoket | The Lightest AI Wristband That Remembers Context
https://memoket.ai/
OpenAI prepares bidirectional voice mode for rollout
https://www.testingcatalog.com/openai-prepares-bidirectional-voice-mode-for-rollout-on-chatgpt/
Grok TTS delivers the most human-like speech”
https://x.com/xai/status/2067654108123910495
In other news, we now have a pipeline for building diffusion draft models (i.e. DFLASH) which are significantly faster than EAGLE/etc. For gemma-4 31b and qwen3.6-27b we saw 30-50% real world decode performance gains. This run is for qwen3-32b (fp8). Seems pretty critical to”
https://x.com/jon_durbin/status/2069876870628155397
OK, I tried GLM-5.2 and this is a good model. Probably the first model good enough to eschew closed models from your workflow entirely (except if you need vision). I know this won’t run on your laptop, but what are the best current vllm/sglang serving recipes?”
https://x.com/gneubig/status/2067936197888930263
Leave a Reply