Just a usual man in an unusual world. Will take note of what I learned from trading Solana memes coin here. Devs | UIX Designer | Photographers
Indonesia
Joined March 2013
- Tweets1.9K
- Following811
- Followers227
- Likes13.4K
sulaksana_ retweeted
Announcing Artificial Analysis Intelligence Index v5, featuring 2 new benchmarks and 3 upgrades. For the first time, the majority of the Index is built on private datasets
The Artificial Analysis Intelligence Index is our composite measure of model capabilities across tasks reflecting real-world usage today. v5 is our biggest upgrade since v4 launched in January. With v5, 60% of the Index is now based on evaluations with private test sets or held-out answers.
Alongside Intelligence Index v5, we’re launching the Artificial Analysis Progress Index v1, which tracks model capability over time on one continuous scale.
New to the Intelligence Index:
➤ FrontierCode 2.0: Cognition's newly-launched private benchmark of real-world coding tasks, graded on whether a repository maintainer would merge the agent's change
➤ Terminal-Bench-Science: 70 agentic scientific research tasks, run in a terminal environment
Upgraded:
➤ HLE-Diamond (replacing Humanity's Last Exam): a refined 1,000-question version of Humanity's Last Exam, split evenly between reasoning and knowledge questions
➤ AA-Omniscience v1.1: An updated dataset with saturated and public set questions removed, new questions added, and ambiguous or erroneous questions revised
➤ CritPt v2: a private, 40-question hard subset of CritPt v2, revised by the CritPt authors and run with a code-execution tool
Staying in the Index: AA-Briefcase, AutomationBench-AA, GDP.pdf and Terminal-Bench 4.0
Removed from the Index: GDPval-AA v2.1, SciCode, and AA-LCR v1.1
Key results:
➤ Claude Opus 5.5 (max) leads at 58, ahead of GPT-6 Astra (max, 57) and GPT-6.1 Sol (max, 55): Anthropic and OpenAI hold the top five models. Gemini 4 Argon (high) is next at 50
➤ The open weights gap widens from 3 points to 12: Kimi K3.1 Pro (weights not yet released) is the top open weights model at 46. Among released weights, DeepSeek V4.1 Flash 1017 (max) and GLM-5.3 (max) lead at 32 - 26 points behind the leader and below Claude Haiku 5.5 (max, 36)
➤ Cost per task varies by more than 13x among the leading models: of the 10 highest-scoring models shown, GPT-6.1 Terra (max) is the cheapest at $0.96 per Intelligence Index task and Claude Fable 5.1 (max) the most expensive at $13.04
sulaksana_ retweeted
Uvdesk is an open source helpdesk software used by universities, companies, and agencies.
Anyone can email their support inbox a bad attachment, the system opens it on its own, and the attacker gets full control of the server.
Zero-click remote code execution.
0click 0auth RCE in uvdesk, found fully autonomously under project apex with @zauthinc.
73 days since disclosure and no response so heres the PSA and POC 🤷♂️
sulaksana_ retweeted
Watch it burn.
“FIRESTARTER” Music Video is out now // riot.com/FIRESTARTER-MV
#KDA #FIRESTARTER #CALLINGALLBLADES
sulaksana_ retweeted
Today only:
If Vector doesn’t find a Medium or higher in your web app, your pentest is free.
Our agents safely attack your site the same way a hacker would.
Spend a few minutes and defend what you’ve built.
The best cybersecurity you’ll find, at a price you’ve never seen.
sulaksana_ retweeted
Public smart contract audits go live this week.
Months of private testing with real clients, and many vulnerabilities caught before they cost users or protocols anything.
As complexity grows on the blockchain, so do paths for failure. Still, most new code is unaudited because audits are slow and expensive.
Not anymore. Zero excuse to forget about cybersecurity.
sulaksana_ retweeted
solana:DNhQZ1CE9qZ2FNrVhsCXwQJ2vZG8ufZkcYakTS5Jpump has lived 10 months in a space where most things don't survive 10 days.
My thesis is simple. AI has revolutionized the world of cybersecurity, and you can bet on this revolution with @zauthinc.
zauth has found vulnerabilities affecting millions of people in the web2 world, alongside hundreds of thousands of dollars in smart contracts exposed to drains.
zauth has hundreds of customers, enterprise contracts, and is building a real company; a startup spawned from a fair launch.
There is nothing like zauth. From the product to the team to the token.
I am not wrong. Just early.
sulaksana_ retweeted
Meet EmbeddingGemma 2, our first natively multimodal open model for on-device embeddings.
It expands beyond text to unify code, images, audio, and video in a shared space. 🧵
sulaksana_ retweeted
Meet Mistral Large 4, aka Le Chonk.
• 1T parameters, natively multimodal. 49B active.
It is the best open weights model from US or Europe on aggregated benchmarks.
• State-of-the-art on critical workloads, including cyber defense, manufacturing and finance and it surpasses closed frontier models on visual grounding.
• Forged in Europe end-to-end and is deployable from Europe via our own Mistral Cloud infrastructure.
• Available to all via API today. Working with cybersecurity partners privately.
Open weights release end of October.
🤖 Made with AI
sulaksana_ retweeted
When pentesting with zauth, you have two options.
a) 50 dollars for a black box test. Our agents attack your web app with nothing but the URL.
b) 150 dollars for a full source code audit.
Nobody does a real pentest for 50-150 dollars. We make cybersecurity accessible.
sulaksana_ retweeted
Our smart contract audit engine found 70+ ETH drainable in a client's live contract.
Right now this engine is private, by contact only, but soon it will be self serve and fully autonomous. Upload the contract, get the audit.
Previously, smart contract auditing was limited to multi-week engagements and tens of thousands of dollars.
We put the most powerful tooling into your hands, for a fraction of the cost and time.
