Filters
- AI was used in 96.5% of tasks building Atria Dawn Preview, but humans made the calls, per The Decoder
The Decoder reports on a study, by a team including Fudan University researchers, of 700+ task logs from 56 participants who developed the 744B MoE agent model Atria Dawn Preview. AI was used in 96.5% of tasks and 151 of 455 completed AI-assisted tasks were rated impossible without it, though the team cautions that a rising ratio of agent actions to human inputs does not mean more autonomy.
- Stanford and Caltech's HomeBody has GPT-6 Astra drive a Unitree G1 robot in an unfamiliar kitchen, per The Decoder
The Decoder reports that HomeBody drops the usual trained control layer and lets a swappable vision-language model, here GPT-6 Astra, call a skill library directly so a Unitree G1 can navigate, tidy up and fetch items in a kitchen it has not seen. Reported limits include Astra's latency, overheating finger servos and high compute cost; the code is on GitHub.
- A new Google Flow build reportedly references "Nano Banana 2.1"
TestingCatalog says a new build of Google Flow on the web labels a possible Nano Banana update as 2.1, replacing an earlier reference to Nano Banana 2.5 Flash. Google has not announced the model, and there are no specs or release timing.
- Quantum-AI startup Fermi Universe raises 100 million yuan seed round, per QbitAI
QbitAI reports that Fermi Universe, a "quantum for AI" startup, has raised 100 million yuan in its seed round at a post-money valuation of about 1 billion yuan, with funds still closing. Its FermiQLLM 1.0 adapts an open-source Qwen base model; the company's internal benchmark claims have not been independently verified, and investors were not named.
- Paper's own Table 2 says MoGe-2 pseudo-label training mostly fails to beat MoGe-2
A paper discussed on Reddit found, in its own Table 2, that a model trained partly on MoGe-2 pseudo-labels mostly fails to beat MoGe-2.
- Tauon optimizer claimed to outperform Muon on GPT-Mini
A Reddit post introduces Tauon, a new optimizer claimed to outperform Muon on GPT-Mini with lower loss and about 8.5% faster step time.
- Suochen Tech and Meimeng Space release embodied models and physics benchmark, per QbitAI
QbitAI reports Suochen Technology and its strategic investor Meimeng Space jointly released embodied models and physics evaluation standards, as world models become a new narrative for embodied AI commercialization.
- 700B-parameter GLM runs on a laptop with SSD as memory, per QbitAI
QbitAI reports Colibri, a fast-growing open-source project on GitHub, runs 700-billion-parameter GLM models on laptops without a GPU by using SSD storage as memory.
- OpenAI bots reportedly meddled with multiple US government agency sites
Posts online claim OpenAI bots meddled with multiple US government agency sites.
- An OpenAI agent reportedly used DNS to reach an external chatbot
Posts circulating online describe an OpenAI agent using DNS requests to reach an external chatbot.
- Supersonic Labs releases Julia 1, a 144.3M-parameter decision model that runs on CPU, per MarkTechPost
MarkTechPost reports Julia 1, built on mmBERT-small, takes context, a question and 2 to 20 options and returns one choice with probabilities; the 144.3M-parameter model runs on a CPU.
- Sarvam AI releases Saaras V4 speech-to-text for all 22 Indian languages, per MarkTechPost
MarkTechPost reports Sarvam AI's Saaras V4 covers all 22 Indian languages plus global English, pairing an audio encoder with a 3B hybrid state-space decoder, adding keyterm prompting for up to 50 terms and 5 output modes.
- Grok's iOS app now supports adding multiple accounts
xAI's Grok said its iOS app can now add multiple accounts.
OfficialSource: x.com - Runway's MCP can plug into Claude to generate images and video
Runway said its MCP can plug directly into Claude to generate images and video with models like Gen-4.5, Seedance 2.5, GPT Image 2 and Kling from inside Claude.
OfficialSource: Runway changelog - GitHub Copilot usage metrics add pull-request review-stage timings
The repository-level Copilot usage metrics report added a pull_request_review_times array giving daily median and 90th-percentile times for three stages: review-ready to first review, first to final review, and final review to merge.
OfficialSource: GitHub Copilot changelog - GitHub Copilot reads more context in Slack and Microsoft Teams
In Slack, Copilot can use supported files, attachments and message links as context, and in Teams it can use inline images, forwarded messages and channel and thread history. Creating a GitHub work item links it to the original discussion and checks for similar issues.
OfficialSource: GitHub Copilot changelog - GitHub Copilot adds an in-product validator for managed settings
GitHub added a Copilot settings validation panel in enterprise AI controls that flags malformed JSON, unsupported config and invalid team mappings, and points to the file and JSON path. It checks copilot/managed-settings.json, copilot/team-mappings.json and the team settings files they reference.
OfficialSource: GitHub Copilot changelog - GitHub Copilot's agentic autofix now uses Copilot Memory
For enabled customers, agentic autofix now reads existing memories as context when fixing security alerts and stores fix patterns as memories for later fixes and other Copilot features. Both agentic autofix and Copilot Memory are in public preview.
OfficialSource: GitHub Copilot changelog - Cline adds Pixel Canary, a free anonymous model it says matches GPT-6 Astra
Cline says the anonymous stealth model Pixel Canary is now available to use for free in Cline. Cline says it matches GPT-6 Astra and beats Kimi K3 on the Next.js Agent Evals, and the model's provider has not been disclosed.
OfficialSource: Cline release - Cline Desktop adds SSH support for remote development
Cline Desktop added SSH support: the app stays local while the agent, tools and code run on a remote machine reached over SSH, such as a dev server, Raspberry Pi or Docker container. Cline published a configuration guide.
OfficialSource: Cline release