Loading activity…
@swarooprm7
RSI @GoogleDeepmind, Opinions mine
AI
Google’s new model targets demanding work; early testing suggests coding promise and flags a case of scope creep.
AI
A post says Carbon scored 88.2% on SEC-bench Pro and is rolling out first to cybersecurity specialists.
AI
A team member says a London role in multi-agent coordination lasts three months, with a possible extension.
AI
The poster says they added Gemini 4 Argon to their AutoResearchExam Pareto curve.
AI
Google DeepMind's expression-of-interest form describes a planned 24-week, in-person NYC project for PhD students starting in early 2027.
AI
Google says Argon is available through its paid API tier and to Pro and Ultra subscribers in Gemini and Search.
AI
Google says the model improves visual design, mask-based editing and subject consistency; the rollout includes Gemini, Flow and AI Studio.
AI
Google announces a 1 million-token output limit, up from 64,000, for its new model. Initial access goes to trusted cyber defenders through the Fairwind Program.
AI
A user testing the three bots says Muse was consistently 2–3 times faster on the same tasks, but even simple requests still took more than five minutes.
AI
AI researchers discuss the relevance of current model evaluation approaches on X.
AI
The user says “gpt-6-astra” gets the same unusual API response as a known internal OpenAI model, while random model names are rejected differently.
AI
New Flash models target agentic workflows and cybersecurity tasks with API and enterprise access.
AI
The method reuses top-layer activations in lower layers during inference on existing models.