BlockBeats reported on September 6th that on-chain analyst Murphy published an analysis titled "Anomaly During the Rise: Whales Didn't Take the Opportunity to Sell." The "BTC On-Chain Cumulative Trend Score" is typically used to measure the direction of whale behavior over the past 30 days, indicating whether it's net buying or net selling. A score close to 1 (black curve on the chart) generally indicates that entities holding thousands or tens of thousands of coins are accumulating. A score close to 0 (yellow curve on the chart) suggests two possibilities: whales are distributing, or whales are remaining inactive.

Looking at the data, the chart curves were yellow when the price surged to $97,000 in January and $82,000 in May, indicating that large investors were net selling during the rebound. This is the standard structure of a bear market rebound, where prices are driven up by short covering and short-term funds, while large investors take the opportunity to sell off their holdings. The upward movement lacks support, and the rebound ends.

The chart curve showing the price rise from $60,000 to $80,000 is black, indicating that whales were net buyers over the past 30 days. This is the first time in the three rebounds that a combination of price increases and large-scale buying has occurred. Of course, large-scale accumulation of shares doesn't necessarily indicate a bottom or a trend reversal. However, it at least shows that this upward move is structurally healthy. Such a structure is typically only seen in the main upward wave of a bull market; that is, price increases coupled with simultaneous large-scale buying.

The bottom of the last bear market didn't have this structure. After the FTX collapse in November 2022, the price was dark around $16,000, but turned yellow during the rally in January 2023. This means large investors bought at the bottom, but took profits during the rally, or at least stopped accumulating. So why did the market rise in January 2023? It was likely driven by three forces: short covering in derivatives, a shift in macroeconomic expectations, and the liquidity crunch following the FTX collapse amplifying price elasticity. The accumulation behavior of small and medium-sized funds is not reflected in this chart. Therefore, by comparison, there is a fundamental difference between January 2023 and August 2026. The price paths are similar, but the on-chain funding structures are different.

Finally, it's worth noting that while this data alone cannot determine whether prices are rising or falling, it can help observe the health of an upward trend. This is very helpful in identifying trends and boosting confidence.

According to Beating AI, Grok Imagine's video creation agent has been upgraded to version 1.5.

This is easily confused with Grok Imagine Video 1.5, which was released in June. Back then, the underlying video model was updated; this update is for the Agent layer. Grok Imagine Video 1.5 was officially launched in June.

The new version incorporates the latest Image 2.0, focusing on improving generated quality, narrative, and multi-shot continuity. Grok claims it excels at connecting multiple shots, making the flow of footage more coherent. It is currently available on the Grok website, iOS, and Android.

In Arena's Text-to-Video rankings, grok-imagine-video-1.5-agent is currently ranked 5th with a score of 1491, higher than Seedance 2.5, Seedance 2.0, and MiniMax H3. However, this is still a preliminary ranking, and the position may change.

According to BlockBeats, on September 6th, Robinhood's token issuance platform, Pons, stated, "Our pace hasn't slowed down. In the past 24 hours, tokens issued on Pons accounted for 73.5% of the total transaction volume across all Robinhood issuance platforms."

According to Dune data, among other token issuance platforms in the Robinhood ecosystem, noxa.fun has a share of approximately 18.2%, followed by long.xyz and pool.trade.