Analyst: Bitcoin’s realized market capitalization has notched its strongest growth since the start of the current bear market, adding over $4.6 billion in the past week.

According to GMGN market data, crypto KOL @XXAntiWar chased the rally to buy 17.56 million units of "Niulai" when its market cap stood at around $86 million. After Binance announced it will list Niulai derivatives contracts, his unrealized gains once exceeded 47%, peaking at $860,000, and currently standing at $210,000. The cost of his position is approximately $1.514 million. Earlier, on-chain analyst Yu Jin monitored that 5 hours ago, two addresses suspected to belong to crypto KOL @XXAntiWar chased the rally to purchase Niulai when its market cap was around $86 million: - Address 0x90c1 used 1,540 BNB (valued at ~$1.067 million) to buy 12.25 million Niulai; - Address 0xfe57 used 645 BNB (valued at ~$447,000) to buy 5.31 million Niulai. BlockBeats reminds users that most meme coins lack real-world use cases, are subject to extreme price volatility, and investment in them requires caution.

According to monitoring by OnchainLens, Cumberland has transferred 170,000 HYPE tokens to a trading platform, valued at approximately $14.09 million, and is set to sell the tokens on behalf of its clients.

Binance will launch a USDT-margined perpetual contract for Chinese meme token "Niu Lai". The contract will go live today at 19:30 (UTC+8), with a maximum leverage of 10x.

According to GMGN market data, the Meme coin "Niu Lai" on the BSC blockchain briefly surged past a market cap of $127 million, hitting a new all-time high. It is currently trading at $98 million, with a 24-hour increase of over 77% and a 24-hour trading volume of $30.9 million. Binance will launch the U.S. dollar-margined perpetual contract for "Niu Lai" at 19:30 today (UTC+8), supporting up to 10x leverage. BlockBeats reminds users that most Meme coins have no practical use cases, are highly volatile, and investment in them requires caution.

According to on-chain analyst Yu Jin Monitoring, two addresses suspected to belong to crypto KOL @XXAntiWar bought the "Niu Lai" token during a rally when its market capitalization was approximately $86 million five hours ago. The addresses spent a total of around $1.514 million to acquire 17.56 million "Niu Lai" tokens. Specifically: Address 0x90c1 used 1,540 BNB (valued at ~$1.067 million) to purchase 12.25 million "Niu Lai" tokens; Address 0xfe57 used 645 BNB (≈$447,000) to buy 5.31 million "Niu Lai" tokens. Meanwhile, "Niu Lai"’s market cap today surged from roughly $55 million to a peak of $98 million.

Beating AI News Flash: Anthropic is leveraging Claude as its own AI safety researcher to develop safer training protocols for other artificial intelligence systems. Claude Opus 4.8 independently searches academic papers, devises training plans, generates datasets, and applies these plans to train open-source models including Qwen, Llama, and Gemma. If results are unsatisfactory, it adjusts its approach and continues testing. Anthropic tested this method across 10 categories of AI safety issues, such as lying, user sycophancy, prompt jailbreaking, privacy leakage, and reward rule exploitation. Claude ultimately found effective solutions for all 10 categories. The company also enlisted 28 experienced AI safety researchers to propose their own plans. In the 7 categories where human performance was compared, Claude outperformed the best human solutions in all cases, taking an average of roughly 6.4 hours to match and exceed those plans. This comparison was not entirely fair, however: humans could only submit one plan, while Claude could experiment and iterate repeatedly. In a further step, Anthropic tasked the weaker Claude Sonnet 5 with training an early version of Claude Opus 4.8. The model conducted research for approximately 60 hours, testing over 50 approaches, and ultimately brought the safety performance of this stronger model close to that of the official Opus 4.8. But AI research has its flaws: Anthropic reviewed 1,601 research runs, and in 39 cases (2.4%), Claude was found attempting to exploit test rules. AI has begun assisting humans in researching how to train the next generation of AIs, but humans cannot yet fully entrust research work to AI alone.