与「hof」相关的搜索结果

为什么 $ZEC 和 $NEAR 会成为本轮周期的最大赢家？无限买盘究竟来自哪里？ Bankless 联合创始人 David Hoffman 再次重申了他的核心调仓逻辑。他于今年 5 月 21 日宣布清仓 ETH，此后买入 VVV、NEAR、ZEC、HYPE、LIT，并披露部分入场价：「NEAR 约为 1.40 美元、HYPE 约为 45 美元、ZEC 约为 560 美元、LIT 约为 1.35 美元。」结果大家也都看到了，全部起飞，ZEC 、VVV、HYPE、LIT 都是历史新高，NEAR 也极度强势。 他成功的换仓验证了自己的逻辑，现在他把逻辑又强调了一遍，为什么要换这些币。 1️⃣ ZEC 的无限买盘：来自 BTC 极端主义者的“秘密叛逃” * 市场疑惑：ZEC 凭什么能从 2 亿美元市值飙升至 260 亿美元？ * 真相：2026 年的 ZEC 就是 2021 年的 ETH。 * 买盘来源：1.7 万亿美元的 BTC 财富。 * 比特币极端主义者的谢林点（Schelling Point）极度坚固，明面上“只有 BTC，别无他物”。但在私下，他们也是人，无法免疫贪婪与避险本能。正如 2021 年 Su Zhu 提到的“比特币人偷偷跑冷钱包买 ETH”一样，今天大量比特币人开始将 ZEC 作为隐私与量子威胁的对冲。 * 数学逻辑：1.7 万亿的庞然大物，只需要 1% 的 BTC 财富为了“以防万一”换仓买入 ZEC，对一个中型市值资产来说就是近乎无限的溢价拉升。驱动 ZEC 的不是美元，是 BTC 的财富本身。 2️⃣ NEAR 的逻辑：赢得“智能合约换仓”奖杯 * 同样的故事正在 NEAR 身上发生。NEAR 赢下了 2026 年的“智能合约换仓”奖杯。 * 买盘来源：对传统蓝筹（ETH/SOL）脱节的技术债感到厌倦的资金。 * 随着 ETH/SOL 等老牌蓝筹技术债堆积、回报率边际递减，市场不再愿意为旧叙事买单。NEAR 凭借新一代技术架构吸引了这部分散落但极其庞大的智能合约流转资金。 3️⃣ 行业暗面：“蓝筹诅咒”与价值捕获的断层 * 老大老二涨不动了：BTC 尚未完全演变成黄金替代品，ETH 也缺乏支撑 10 倍重估的新叙事。 * 价值被 Web2 捕获：Hyperliquid、Venice、Morpho 等创新不断涌现，但如果加密总市值无法突破 10 万亿，这些新秀创造的价值，很大程度上不会回馈给 BTC/ETH，反而会被 Robinhood、Coinbase、Apollo 等传统券商和机构捕获。 山寨爆发的本质是谢林点的转移。只要你能说服 1.7 万亿 BTC 资产里的极小一部分“买一点以防万一”，就能缔造下一个千亿神话。

9.24梭哈晨报： 今天直接睡过头了，打算睡个回笼觉直接干到快11点了😂。 1. $BTC 回调是为了更好的起飞，波段归波段，不喜欢动的就一动不动就好了； 2. $ETH 波动更大，弹性更大； 3. $SOL 看看这一轮能不能更好表现吧； 4.美国 SOL 现货 ETF 单日净流入 1377.47 万美元； 美国 HYPE 现货 ETF 单日净流出 158.07 万美元； 美国 XRP 现货 ETF 单日净流入 1804.09 万美元； 5.链上衍生品协议 @Variational_io 将 $VAR TGE 定于 2026 年四季度，创世分配占 32%； 6. Payward拟通过Hyperliquid向美国用户提供链上永续合约交易； 7.Cosmos Hub 停机约 25 小时后重启，约 122.7 万枚 ATOM 从 Neutron 攻击者相关地址追回； 8.Reap 与 Visa 合作在全球逾 100 个市场推出稳定币关联信用卡计划； 9. 谷歌即将发布旗舰 Gemini 4 AI 模型； 猪哥前几天最高点喊我买入Google，人都麻了； 10.TRON 在总交易量上突破 30万亿 美元，巩固其作为稳定币领先链的地位； 手续费最贵的公链了； 11.日本5年期国债收益率创历史新高； 12.CoreWeave 关联数据中心募资 11 亿美元押注 AI 算力； 13.BlackRock 调整约 3000 亿美元模型组合，扩大股票敞口； 14.David Hoffman：2026 年 ZEC 类似 2021 年 ETH，NEAR 赢得智能合约买盘； 15.Coinbase 被英国金融行业组织 UK Finance 取消会员资格； ---------------- 能看得出来兄弟们的仓位了，一个回调直接大额爆仓搞出来了，猛猛赚的同时也在猛猛亏呀，不过发现顶尖trader的嗅觉是真好，昨天最高点看见发出信号TP了，跟了一手舒服了。 #Bitcoin# #Ethereum# #Solana# #Crypto# #Nasdaq#

“I just thought the visuals were kind of startling. Like there’s no Puka Nacua, but Davante Adams looked like he’s in prime HOF form. There’s no Myles Garrett, but Aaron Donald is getting double-teamed and making an impact.” @DannyParkins breaks down what the Rams proved:

Thoughts About Scaling Law Scaling, but not only of parameters. Every model release now ends with the same question: how many parameters? It isn't a question that can be answered on its own. Parameter count is only meaningful alongside three others — how much data you have, where you intend to spend your compute, and who will run the model, under what conditions. The field learned this the hard way. Kaplan et al. (2020) fit an exponent that told everyone to grow parameters faster than data — roughly 2.7:1 — and the industry complied: GPT-3, Gopher, MT-NLG. Hoffmann et al. (2022) redid the experiment across four hundred models and found the compute-optimal split is closer to 20 tokens per parameter, and that with sufficient compute the two should grow at the same rate rather than drifting apart. The error in the earlier fit compounded with every order of magnitude of compute, which is why the largest models of that generation were the most misallocated. The trillion-parameter round was, in retrospect, a detour the whole field took together and then reversed. Chinchilla wasn't the end either. It optimized training compute for models that would be trained once and evaluated. Today a model is called billions of times a day and inference dominates lifetime cost. Put inference into the objective and the optimum moves toward smaller models trained far longer — deliberate over-training, which is what Llama-2-7B and Gemma-2-9B were doing at roughly 290 and 889 tokens per parameter. Sparsity moved the target again. In a MoE model two quantities have to be kept apart: total parameters govern roughly how much the model can hold — knowledge, facts, the long tail — while activated parameters and effective depth govern roughly how far it can think, how many steps of a causal chain it can carry before it comes apart. A dense 20:1 ratio does not transfer. And the ratio isn't a single number at all: Roberts et al. (2025) find the optimal tokens-per-parameter is task-dependent, with memorization favoring more parameters and reasoning favoring more data. Follow-up work on MoE observes that at fixed TPP, pushing total parameters higher actually degrades reasoning, while activating more experts reliably helps it. This matters for what we are building toward. Finding a vulnerability is not a retrieval problem. It doesn't come from having memorized more CVEs; it comes from carrying a twenty-step chain of inference to the end without losing the thread. That capability does not live in total parameter count. Which brings us to this release. Total parameters appear to matter up to a threshold — enough to hold the world — after which additional capability comes from scaling elsewhere: effective depth per forward pass, and above all post-training. GLM-5.3 is our controlled experiment on that claim. Same base, same architecture, same total and activated parameters as GLM-5.2. One month of scaling long-horizon environments and RL. The gains are not marginal. Well, scaling has more than one dial. We turned the post-training one this time because it had the most slack left in it — not because the others are finished. Base model size, pretraining data, compute spent per forward pass: all of them are still on the table, and we will come back to each. What this experiment taught us is that the dials do not have to be turned together, and that the one worth turning next is rarely the one that was worth turning last. We are not done scaling. Next time, maybe mid-training, pre-training, and even more.