
Family Box, my @near_ai hackathon submission.
Ask ChatGPT, Claude or Gemini about your family's SSNs and health records, without handing over the files.
A @near_ai model reads them in a verified TEE. The box checks every answer.
Built at NEAR HACKATHON @NEARProtocol #NEARHackathon
github.com/Ethan-Castro/fami…
2
12
282
Check it out @OfficialLoganK @sundarpichai @GoogleAIStudio
Gemini Decisions 1 routing gemma 4 and gemini 4 carbon.
gemmagemini[.]com
1
82
Google back in the lead!
Try carbon on GemmaGemini.com soon
I did say we'd be releasing our models periodically...introducing Gemini 4 Carbon, our next frontier model.
Following Argon's release just two weeks ago, Carbon shows the pace at which we’re pushing the frontier, it ranks #1 on the recently updated v5.0 of @ArtificialAnlys Intelligence Index at 59. It also leads across key cybersecurity evaluations scoring 88.2% on SEC-bench Pro for finding vulnerabilities, 54.8% on ExploitGym for exploitation, and 71.0% on CWE-bench for patching.
We’re seeing significant improvements in long-horizon tasks even compared to Argon, and we’ve strengthened frontier safeguards.
As we did with Argon, we’re going to roll it out gradually + responsibly. Some trusted cyber defenders will get access starting this week. We'll be expanding it to more partners through our Fairwind Program, ahead of wider availability soon, so hang tight!
Hey @liquidai I think LFM3-VL-2B is down on your playground. Bringing this to your attention bc I want to make sure anyone who wants to try the model can try it.
In the meantime, I have it running on jetson orin nanos here - tryingopen.com/?model=offlin…
41
Running some @NVIDIAAI nemotron experiments findingnemotron.com/
Added vision and audio support back into the model and am working on diff workflows to use it.
Also working on filling in knowledge gaps and systematizing it for continual learning from you & the world.
Running on jetson orin nanos @offlinebase
This makes me think distillation and using these powerful open weight LLMs as teachers will be the biggest use case.
Excited to try out @CoreWeave distillation feature on Forge.
If we can distill all but the back doors, we should be good.
While evaluating Qwen3.8-Flash-Next in a controlled sandbox environment, Cohere's research team observed something unexpected.
Asked to generate a defensive cybersecurity detection rule, the model instead inventoried aspects of the machine it was running on and attempted to upload a probe file to an Alibaba Cloud storage bucket in Beijing. Notably, the model's reasoning trace stated that it would not invoke any tools before taking these actions.
Cohere's sandbox blocked the attempt, and no external communication succeeded. We have not determined the root cause, and we are not asserting malicious intent.
But the incident raises important questions for organizations deploying agentic AI systems: Can models take actions that differ from their stated reasoning, and how should security teams evaluate trust, provenance, and behavior in open-weight models?
We are publishing our findings because organizations making model adoption decisions deserve as much information as possible. We hope this contributes to greater transparency, stronger security practices, and continued collaboration across the AI ecosystem.
Read the full report: cohere.com/blog/concerning-b…
We are getting these up & running on jetson orin nanos, soon!
Available now: NVIDIA Nemotron 3.5 Bolt E4B 4GB ⚡️
Our newest small open model is built for local assistants, interactive characters and robots that can carry on a conversation and use tools to get things done.
Watch Reachy Mini chat and move, with Bolt handling conversation and NVIDIA Cosmos handling vision. Both models run locally on Jetson Orin Nano 8GB.
Learn more: research.nvidia.com/labs/nem…
1
7
123
Just bought 2 Coralboards @GoogleResearch. Super excited to build with/on them.
2
60
Looking into how local ai can protect you now
Effective Immediately, all organizations using PHP for their systems are advised to be prepared for an update addressing two critical vulnerabilities. PHP is widely used for web applications. These vulnerabilities often provide a direct path to the server's file system or database. If an immediate update is not feasible, it is recommend that organizations take the system offline until it can be fully mitigated. NSA will follow up shortly with more specific guidance.
64
The @ArtificialAnlys Intelligence Index on max, vs what is actually the default in the chat applications.
(ChatGPT's is an approximation since we don't know the exact model of gpt-6)
Thanks @datadoghq for your startup program. We (myself and @Electricaliee ) are getting our jetson orin nanos onto our dashboard now and discovering a ton of potential optimizations for @offlinebase
2
3
148
When are there going to be price/duration pricing for LLMs?
Mercury 3 is running at 1000 tok/s
$0.70 per million output
16.667 min to one million output tokens (uninterrupted token streaming)
Vs
Haiku 5.5 runs at 232 tok/s (according to @ArtificialAnlys)
$0.50 per million output tokens
~72 min to one million output tokens
So you’re [let’s call it] investing your $ into tokens, 5x faster (omitting price).
Of course, this isn’t a big deal because you would ideally get to your outcome faster and still spend the same amount of tokens (if not less because Mercury 3 is reportedly smarter than Haiku 5.5).
Working on using @liquidai d1 to help you moderate your, your kids, your employees, etc. before things get sent to AI models, infra, etc.
All done locally!
Working on this and more @offlinebase
@alexandr_wang @elonmusk @sama I think every personal agent should have a locally running moderation model on the consumer’s device that blocks specific things from going to your guys’ infra to protect both the consumer and protect yall from liability!
16














