Binance Square
橙子Joyce
877 Posts

橙子Joyce

价值投资者:以十年为单位投资美股及BTC.ETH.BNB.SOL.推特X:@Joyce88ai
BNB Holder
BNB Holder
Frequent Trader
8.6 Years
629 Following
28.5K+ Followers
31.0K+ Liked
Posts
PINNED
·
--
Verified
Article
No matter whether open-source AI wins or closed-source AI wins, cybersecurity will be the ultimate winner! Palo Alto and CrowdStrike seize the windfall of the AI inference era!The Wall Street financial giant Wells Fargo Bank recently released a research report stating that cybersecurity supergiant $CrowdStrike (CRWD.US)$ and $Palo Alto Networks (PANW.US)$ are seeing momentum in growth of orders for cybersecurity software products driven by the rapid expansion of the AI inference market. After AI moves into the stage of large-scale, massive inference and AI intelligent agent-style workflows (Agentic AI), cybersecurity demand is not simply likely to accelerate growth by “updating and iterating alongside cutting-edge AI technologies.” Instead, it may see a structural incremental expansion far beyond traditional IT spending.

No matter whether open-source AI wins or closed-source AI wins, cybersecurity will be the ultimate winner! Palo Alto and CrowdStrike seize the windfall of the AI inference era!

The Wall Street financial giant Wells Fargo Bank recently released a research report stating that cybersecurity supergiant $CrowdStrike (CRWD.US)$ and $Palo Alto Networks (PANW.US)$ are seeing momentum in growth of orders for cybersecurity software products driven by the rapid expansion of the AI inference market. After AI moves into the stage of large-scale, massive inference and AI intelligent agent-style workflows (Agentic AI), cybersecurity demand is not simply likely to accelerate growth by “updating and iterating alongside cutting-edge AI technologies.” Instead, it may see a structural incremental expansion far beyond traditional IT spending.
PANWUS+0.26%
CRWDUS+0.16%
ZSUS+0.30%
PINNED
Article
An Apple research team ran a reinforcement learning experiment with 9 models and 11 languages to find out: if the training data is only in one language, can the model’s problem-solving abilities be applied to questions in other languages?The answer is yes, and the effect is quite noticeable. Training with only one language still makes the model stronger across many other languages as well. For example, on a French test, training directly in French improved the average score by 25.6 percentage points. If you don’t train in French at all and only use Spanish questions for practice, you can still improve the score on the French test by 24.6 percentage points—just a 1-point difference. What the model learns isn’t only problem-solving for a single language; it also learns some problem-solving methods that can be transferred and continued to be used in other languages. So in the future, if you want to strengthen the model’s Chinese reasoning, you may not need to remake all reinforcement learning data into Chinese.

An Apple research team ran a reinforcement learning experiment with 9 models and 11 languages to find out: if the training data is only in one language, can the model’s problem-solving abilities be applied to questions in other languages?

The answer is yes, and the effect is quite noticeable. Training with only one language still makes the model stronger across many other languages as well.
For example, on a French test, training directly in French improved the average score by 25.6 percentage points. If you don’t train in French at all and only use Spanish questions for practice, you can still improve the score on the French test by 24.6 percentage points—just a 1-point difference.
What the model learns isn’t only problem-solving for a single language; it also learns some problem-solving methods that can be transferred and continued to be used in other languages. So in the future, if you want to strengthen the model’s Chinese reasoning, you may not need to remake all reinforcement learning data into Chinese.
avatar
@奋斗1688
is speaking
[LIVE] 🎙️ Understand Binance’s Six Major Sectors and See BNB’s Long-Term Value
4.8k listens
橙子Joyce
·
--
No matter whether open-source AI wins or closed-source AI wins, cybersecurity will be the ultimate winner! Palo Alto and CrowdStrike seize the windfall of the AI inference era!
The Wall Street financial giant Wells Fargo Bank recently released a research report stating that cybersecurity supergiant $CrowdStrike (CRWD.US)$ and $Palo Alto Networks (PANW.US)$ are seeing momentum in growth of orders for cybersecurity software products driven by the rapid expansion of the AI inference market. After AI moves into the stage of large-scale, massive inference and AI intelligent agent-style workflows (Agentic AI), cybersecurity demand is not simply likely to accelerate growth by “updating and iterating alongside cutting-edge AI technologies.” Instead, it may see a structural incremental expansion far beyond traditional IT spending.
橙子Joyce
·
--
An Apple research team ran a reinforcement learning experiment with 9 models and 11 languages to find out: if the training data is only in one language, can the model’s problem-solving abilities be applied to questions in other languages?
The answer is yes, and the effect is quite noticeable. Training with only one language still makes the model stronger across many other languages as well.

For example, on a French test, training directly in French improved the average score by 25.6 percentage points. If you don’t train in French at all and only use Spanish questions for practice, you can still improve the score on the French test by 24.6 percentage points—just a 1-point difference.

What the model learns isn’t only problem-solving for a single language; it also learns some problem-solving methods that can be transferred and continued to be used in other languages. So in the future, if you want to strengthen the model’s Chinese reasoning, you may not need to remake all reinforcement learning data into Chinese.
阿波罗1111
·
--
May every step you take today be steady and bright.
愿你今天的每一步都走得安稳、明亮。
Lily雪莉呀
·
--
Take what can’t be carried away—what can’t be lost. Let the heavy rain erode it…
大仁Jaron
·
--
Nvidia invests $1.5 billion to take a stake in SB Energy, targeting 8GW of AI computing capacity; OpenAI will be the sole tenant
On August 17, Nvidia announced a strategic partnership with SB Energy to provide credit guarantees for the PORTS-Pike Technology Park in Ohio, helping build it into a massive data center dedicated to Nvidia’s AI computing infrastructure, with OpenAI as the sole tenant. At the same time, Nvidia invested $1.5 billion in SB Energy.
With this move, Nvidia directly entered the underlying infrastructure segment of land, power, and facilities (LPS). Its significance goes far beyond a typical data center investment. The key change is that Nvidia is using its own creditworthiness to “underwrite” financing for AI infrastructure, in return for long-term and exclusive allocation of computing capacity. This marks the chip giant expanding from being merely a supplier of AI hardware into a core participant and organizer in building and operating “AI factory” infrastructure.
英鸿³³₇
·
--
Ather's pinnacle masterpiece 01, Stage 2 (No. 18) — Binance AMA is something we hope everyone will join. The featured live stream has airdrop surprises too ❤️❤️❤️
路人1688-luren
·
--
There are many beautiful landscapes. The road ahead will be smooth, with wealth and prosperity close by. Let’s encourage each other.
晚风 Vesper_1688
·
--
🍵A cup of clear tea—refine your mind and keep your composure ☕
Invest like tasting tea: don’t rush, don’t be restless 🌱
Let your heart settle 🧘, hold the rhythm steady, and don’t push for quick results 🌿
Look past the market’s ups and downs 📊, and quietly wait for the cycle to arrive ⏳
Time will be the answer, as it deepens and settles 💫

$BNB 🧧🧧🧧

#中国7月产出零售投资全线不及预期

#1688家族family
从0开始1688
·
--
📌Oh my, is Ethereum about to climb up 15,000? 😱

On August 17, BitMine’s chairman Tom Lee cited technical analysis, expressing expectations for a potential technical breakout in Ethereum.

Ethereum’s current price is only 3.5% away from the top boundary of the Ichimoku cloud of the daily chart—this would be the first time it has closed above that key resistance since October 9, 2025.

At present, the price of Ethereum is around $1,906, with cloud resistance just within reach.

Tom Lee continued his earlier bullish view on Ethereum. He previously posted publicly in July that the ETH/BTC ratio would strengthen in the second half of 2026. His core logic centered on the combined momentum of ETH’s monetary narrative, stablecoin growth, and the tokenization of assets.

The top of the Ichimoku cloud is regarded by technical analysts as an important reference for trend reversals. If it breaks out effectively, it may open up further upside room.
从0开始1688
·
--
Meme potential is limitless “Niu Lai” Web 3 and Web 2 hand in hand, charging forward!
My friend said, “Watching it for 80 minutes, I regretted it! Not watching it—regret it for a lifetime!”
“I laughed for an hour after watching for just one hour.”
On August 17, the production company’s location in Dalian previously removed the film from schedules nationwide due to concerns about “impacting the city’s image,” but today reports say that some local theaters have started reinstating screenings. The earlier nationwide rumor that “from August 17, no new screenings will be added” has not been officially confirmed as a unified notice; most cities are still screening normally.
The total box office of the animated film (Niu Lai) predicted by Maoyan Professional Edition is about 135 million yuan. At the time of the prediction release, the film’s cumulative comprehensive box office (including pre-sales) has already exceeded 17 million yuan and is nearly catching up to the film’s predicted total box office given by Maoyan three days ago. If the total box office of (Niu Lai) reaches 135 million, the film will enter the top 35 of the 2026 China national annual box office rankings. The films it overtakes include major commercial releases such as (Zhua Te Wu), (Xing He Ru Meng), and (Yong Wu Zhi Di).
凯哥的进击
·
--
Bearish
🎁🎁Reply today to claim the red envelope 🎁🎁$SOL
✅Use this method to get the results ✅
🌹Thank you for helping share and repost THS 🌹
This image is a screenshot from the plaza fans.
BANK大鹤
·
--
Toyota and BYD can actually use USDT to buy cars. #USDT🔥🔥🔥
In Bolivia, Toyota and BYD dealerships have already started supporting USDT payments for car purchases. Since the local foreign exchange reserves are insufficient and the local currency has depreciated by nearly 50%, USDT has become an important option for businesses and individuals to hedge against currency exchange-rate risk.
大丽7613
·
--
$BTC

On Wednesday, there’s an important announcement.
Pay attention to the trajectory of the CLEAR Act.
Trump himself will personally attend the White House Crypto Summit, and the heads of two key regulators, the SEC and the CFTC, will also all be there.
周周1688
·
--
🚀🚀August 17, 2026|Crypto Daily
$BNB 🧧🧧

ETF fund outflows: safety incidents happening frequently
A new week begins with a modest rebound.

① Bitcoin ETF: outflows remain a pressure point
From August 10–14, the U.S. spot Bitcoin ETFs saw net outflows of about $390M last week, becoming one of the largest weeks of capital withdrawal in recent times.
The key question the market is focused on this week is very simple:
Will ETF fund outflows start to reverse?
② SafePal: data of about 39,800 users affected
SafePal confirmed that its order-tracking plugin has a security vulnerability, resulting in unauthorized access to some users’ information such as names, email addresses, phone numbers, and addresses.
Notably:
Seed phrases, private keys, and asset information were not affected.
③ Bits of Gold: data leak affecting about 200,000 customers
Israel’s large regulated crypto broker Bits of Gold also experienced a data leak.
With both incidents occurring at the same time, it once again reminds the market:
Crypto security issues are not only about wallets and private keys—user data itself is also a target for attacks.
④ The first U.S. spot Bitcoin ETF is set to close
Hashdex DEFI will have its final trading day today, then move into liquidation.
This is the first spot Bitcoin ETF in the U.S. to fully close, and it also reflects that competition in the ETF market is intensifying—capital will not flow evenly to all products.
⑤ What’s truly worth watching this week
The Wyoming Blockchain Symposium was held August 17–20.
Also this week:
• FOMC Minutes
• White House meeting with Crypto industry CEOs
• CFTC Innovation Advisory Committee’s first meeting
• Continued attention on UBS and Paul Tudor Jones’ BTC ETF allocation
In addition, Cboe is moving forward with leveraged products such as the 3× Bitcoin / Ethereum ETF.
🚀 Top gainers today
DOLOUSDC: +28.9%
DOLOUSDT: +27.5%
Small-cap assets continue to show clear bursts of capital activity, but in the current low-liquidity environment, short-term volatility risks are also higher.
One-sentence summary:
On Monday, BTC only rebounded slightly—what really matters to watch is ETF fund flows, security incidents, and the policy and macro signals coming this week.
The market looks calm,
but this week’s calendar isn’t.
#1688家族family
$BTC $ETH
必按666 势不可挡
·
--
#PREDICT Predicting the market: Binance’s new platform—smooth, fluid, and secure. Have you experienced it? It’s really awesome!

#美国拟迫各国在美中AI阵营选边
橙子Joyce
·
--
No matter whether open-source AI wins or closed-source AI wins, cybersecurity will be the ultimate winner! Palo Alto and CrowdStrike seize the windfall of the AI inference era!
The Wall Street financial giant Wells Fargo Bank recently released a research report stating that cybersecurity supergiant $CrowdStrike (CRWD.US)$ and $Palo Alto Networks (PANW.US)$ are seeing momentum in growth of orders for cybersecurity software products driven by the rapid expansion of the AI inference market. After AI moves into the stage of large-scale, massive inference and AI intelligent agent-style workflows (Agentic AI), cybersecurity demand is not simply likely to accelerate growth by “updating and iterating alongside cutting-edge AI technologies.” Instead, it may see a structural incremental expansion far beyond traditional IT spending.
橙子Joyce
·
--
An Apple research team ran a reinforcement learning experiment with 9 models and 11 languages to find out: if the training data is only in one language, can the model’s problem-solving abilities be applied to questions in other languages?
The answer is yes, and the effect is quite noticeable. Training with only one language still makes the model stronger across many other languages as well.

For example, on a French test, training directly in French improved the average score by 25.6 percentage points. If you don’t train in French at all and only use Spanish questions for practice, you can still improve the score on the French test by 24.6 percentage points—just a 1-point difference.

What the model learns isn’t only problem-solving for a single language; it also learns some problem-solving methods that can be transferred and continued to be used in other languages. So in the future, if you want to strengthen the model’s Chinese reasoning, you may not need to remake all reinforcement learning data into Chinese.
Log in to explore more content
Join global crypto users on Binance Square
⚡️ Get latest and useful information about crypto.
💬 Trusted by the world’s largest crypto exchange.
👍 Discover real insights from verified creators.
Email / Phone number
Sitemap
Cookie Preferences
Platform T&Cs