GPT 6 Astra (high) scores 92.9% on WeirdML and matches Fable 5.1 (max) for the top score.
It sets a new individual high score on 6 of the 17 tasks, and is overall very solid.
Interestingly it writes significantly less code than Sol and other recent GPT models, although still more than Claude.
I will do a more careful analysis when the max and pro max runs are done.
Posted by Håvard Ihle, 3 hours ago
It sets a new individual high score on 6 of the 17 tasks, and is overall very solid.
Interestingly it writes significantly less code than Sol and other recent GPT models, although still more than Claude.
I will do a more careful analysis when the max and pro max runs are done.
Posted by Håvard Ihle, 3 hours ago
This media is not supported in your browser
VIEW IN TELEGRAM
From GPT-4 to GPT-6.
Sparks was a remarkably prescient paper that got a lot of pushback at the time, but absolutely sensed where the vibes were heading with LLMs based on a lot of qualitative experiments. It deserves credit in retrospect.
Posted by Ethan Mollick, 29 minutes ago
https://www.microsoft.com/en-us/research/publication/sparks-of-artificial-general-intelligence-early-experiments-with-gpt-4/
Sparks was a remarkably prescient paper that got a lot of pushback at the time, but absolutely sensed where the vibes were heading with LLMs based on a lot of qualitative experiments. It deserves credit in retrospect.
Posted by Ethan Mollick, 29 minutes ago
https://www.microsoft.com/en-us/research/publication/sparks-of-artificial-general-intelligence-early-experiments-with-gpt-4/
Gemini 3 achieves state of the art performance in SpatialBench. A Spatial reasoning benchmark for VLM to test their tracing, 3d visualization ability, and reasoning across each.
Posted by spicylemonade, 9 months ago
Posted by spicylemonade, 9 months ago
I havent updated this benchmark in a while. Astra completely saturates my spatial reasoning eval. I am at a loss for words, and i'm declaring LLM vision solved. Every so often when a new model came out I would test it on a sample question and it failed. I tested Astra and it kept getting answers correct, so I evaluated it...
https://x.com/spicey_lemonade/status/1991744517528252686?s=20
Posted by spicylemonade, 1 hour ago
https://x.com/spicey_lemonade/status/1991744517528252686?s=20
Posted by spicylemonade, 1 hour ago
Do It by Code
Photo
It seems like they trained it to know precise coordinates in reasoning (without any external tool calls, at least not on the client-side)
Heads up, there is a 1-day full MikroTik RCE chain against SSH being used in the wild. Patch was released yesterday, so if you have a MikroTik router with ssh open on the internet, it may already be compromised. Detection guidance and IOCs courtesy of @CERT_Polska_en https://cert.pl/en/posts/2026/09/vulnerabilities-in-mikrotik-routeros-actively-exploited/
Posted by Costin Raiu, 10 hours ago
#rce #ssh
Posted by Costin Raiu, 10 hours ago
#rce #ssh
Media is too big
VIEW IN TELEGRAM
One shotted with GPT6 Astra
I now have a fork of UEVR where I can play any UE game within a volumetric window!
This actually renders the game in 3D - meaning, two real game cameras, not some depth map bullshit. Everything looks right
This is an alternative (and more comfortable IMO) way to play most 2D->VR hacked games as there’s no motion sickness, and no need to redesign the UI and interaction systems
Needs a lot more polish before I publish it but OMG IT WORKS!
Posted by Azad Balabanian, 7 hours ago
I now have a fork of UEVR where I can play any UE game within a volumetric window!
This actually renders the game in 3D - meaning, two real game cameras, not some depth map bullshit. Everything looks right
This is an alternative (and more comfortable IMO) way to play most 2D->VR hacked games as there’s no motion sickness, and no need to redesign the UI and interaction systems
Needs a lot more polish before I publish it but OMG IT WORKS!
Posted by Azad Balabanian, 7 hours ago
A similar project: elliotttate/UEVR-6DOF-Window
Government Rails Site Hit Hours After CVE Patch
Article, Comments
Article, Comments
TL;DR: A critical Ruby on Rails ActiveStorage RCE vulnerability was patched on a government site late at night. Just eight hours later, attackers tried exploiting it; likely using a public proof of concept released within hours of the fix. The patch worked, so the site was not compromised, despite continued probing throughout August. Main lesson: critical security updates must be deployed within hours, because attackers can reverse-engineer patches almost immediately
idk how they RL'd Astra but it seems to LOVE writing markdown documents everywhere.
this reminds me of the Opus 4.5 or even the Sonnet 4.5 era almost a year ago where they made (slop) markdown docs for fucking everything. not a fan of this one.
Posted by Josh, 10 hours ago
this reminds me of the Opus 4.5 or even the Sonnet 4.5 era almost a year ago where they made (slop) markdown docs for fucking everything. not a fan of this one.
Posted by Josh, 10 hours ago
especially if you use the trigger word "investigate"
it will write a REPORT
enjoy your 300 lines of markdown describing why the telegram macos app is using so much ram right now
We are aware of a security incident on @Liquid_BTC. Purported white-hat hackers have withdrawn ~4,000 BTC (~$320 million) from the Liquid Federation wallet. The @Blockstream team is working on contacting them on-chain with a signed message.
What we know so far is that the funds were withdrawn via the SideSwap PAK (Peg-out Authorization Key), but that key was not compromised, nor were any others.
Exchanges have been notified and have already paused (or will pause) LBTC deposits and withdrawals. Other Liquid assets such as USDT, DePix, and RWAs are unaffected by this security incident.
Bridge nodes have been temporarily disabled, so no new transactions can be submitted to the network. Effectively, the Liquid sidechain is paused until this issue is resolved.
Liquid wallets will be impacted, and we're sorry for any inconvenience. Federation members are actively working on resolving this so we can restore normal network activity.
You can monitor the situation via @mempool's http://liquid.network site below:
https://mempool.space/address/bc1qdlld6antmv4xug242ed83q7k4rqw50cwfns38szx4qu2f4jwaxxsuhwxxr
Posted by Liquid Network 🌊, 3 hours ago
Re: LLMs as a Cognitive Virus
I love this line of thinking but the framing comes across to me as a bit inflammatory and uncharitable. Just about anything involved in the exchanged of ideas can be viewed as a virus! That is because, when viewed from an evolutionary biology perspective, ideas are the cognitive equivalent of genes (in fact, this is where the term meme comes from).
There's a whole field, evolutionary memetics, dedicated to this. Cultural values, marketing, religion, social media, education, propaganda, etc. can all be viewed quite naturally through this lense. Even the terms we use in every day language (eg, going "viral") seem to intuitively grasp this.
Is my friend a virus for recommending me their favorite book? Well, yes I guess, but that's kind of just how ideas work. Just my two cents
Murfalo, 22 hours ago
I love this line of thinking but the framing comes across to me as a bit inflammatory and uncharitable. Just about anything involved in the exchanged of ideas can be viewed as a virus! That is because, when viewed from an evolutionary biology perspective, ideas are the cognitive equivalent of genes (in fact, this is where the term meme comes from).
There's a whole field, evolutionary memetics, dedicated to this. Cultural values, marketing, religion, social media, education, propaganda, etc. can all be viewed quite naturally through this lense. Even the terms we use in every day language (eg, going "viral") seem to intuitively grasp this.
Is my friend a virus for recommending me their favorite book? Well, yes I guess, but that's kind of just how ideas work. Just my two cents
Murfalo, 22 hours ago
I've factored the RSA keys of a Certificate Authority from the 90s
comments
The author explores the security of RSA cryptography, noting that small keys can be factored on a modest desktop computer and some old root certificates are still vulnerable.
comments