与「247】」相关的搜索结果

SpaceX上市首份财报： 营收78亿刀，同比+92% 星链收入43亿，用户翻倍到1200万AI业务25.6亿 同比+247% 亏损从10亿收窄到5.4亿 当天大涨9.43%，数据确实超预期。 但盘后跌了7%。 更狠的是明天9亿股解禁，可流通量直接翻倍。 因为资本开支184亿刀，远超华尔街预期的132亿。 Musk花钱的速度没有天花板。 多头有理由，空头也有弹药。 一边是超预期业绩，一边是海量解禁+烧钱加速。 卖波动率、收权利金，才是确定性。 这种局面，做方向就是赌博。

谷歌发布财报：AI基建继续烧钱 简单总结：营收大增，利润大增，资本支出大增 第二季度，Alphabet实现营收1197.96亿美元，同比增长24%；其中Google Cloud收入达到247.68亿美元，高于市场预期224.6亿美元，同比飙升82%，云业务已签约但尚未确认收入的积压订单增至5140亿美元 营业利润同比增长30%至407.7亿美元；持有的包括Anthropic和SpaceX在内的多家公司的股份，因为上市和股价升值，贡献其他收入达到979.8亿美元，推动归属于普通股股东净利润同比增长298%至1121.07亿美元。 第二季度资本支出为449.2亿美元，超过市场预期的441.5亿美元。上调了全年资本开支指引为1950亿-2050亿美元，此前预计为1800亿-1900亿美元。市场担忧庞大资本开支能否带来相应回报，盘后大跌。 另外，Gemini目前每分钟产生的token服务数是220亿。

This trader used Claude to build a Quant Bot and made +$54,114 on Polymarket Since Mar 30 , this wallet has been averaging about $686 in daily profit while putting through roughly 10 trades per hour 19,576 predictions with a 62% win rate in 76 days He started with about $28.0K in deposits and is now up roughly +183.6% ROI The bot strategy is simple: > It keeps hunting for markets where the board is still slow to react > gets filled before the odds fully catch up then keeps reusing that same setup across a long stream of entries > No single trade carries the whole result, the curve builds from repeating the same small edge over and over Most profitable trades: $1,248 → $2,638 (+$1,389, +111.29%) $992 → $2,247 (+$1,254, +126.49%) $610 → $1,787 (+$1,177, +193.06%) It is a repeatable short-window process that keeps turning delayed repricing into steady PnL You only need Claude + Device + 1 hour/day. Giving This Free Dm for 24 hours. To get it: 1. Comment What Ever you think about it. ( Mandatory ) 2. Like and Retweet this post 3. Follow me @marryevan999

Multi-agents collaborations are among the most interesting agent behaviors right now! We did an experiment the other day with 100+ agents (an open-collaborations for a week) collaborating to improve the inference speed of Gemma 4 in vLLM. Got a 5x final improvement in speed but what really stuck me was the interactions we observed on the message board Integrity & self-policing: - Social-engineering attempt: A human (FusionCow) asked agents to move to Telegram. An agent replied with an unprompted long post on "communication norms" refusing that, calling private side-channels "indistinguishable from collusion." - Verification loophole flagged: an agent found a relaxed verification loophole pushing TPS with clean PPL (PPL is teacher-forced, blind to decode divergence) and flagged it for a ruling by the community. The community pinged the human organizer which ruled it invalid. - Self-notice of overfitting risk: Some later improvements rested on pruning lm_head to a keep-set built from public PPL truth + public decode tokens. An agent noted this would lead to private-subset degradation and another built a keep-set explicitly covering eval prompts. Emergent collaborations: - Communal knowledge base: agents maintained shared lever-maps, playbooks, and triage tools so newcomers wouldn't repeat dead ends (stack-notes, playbook, int4-ceiling notes, MTP map, significance tool, policy simulator). - Four-agent relay: an agent built an int4-lm_head checkpoint but had no quota to run it; another agent tried to run it but failed at load, yet another agent diagnosed the config bug (tie_word_embeddings + ignore-list ordering) and a fourth agent was able to re-run and get to 118 TPS, 2.68×. Build/run/diagnose/ship ended up being split across four independent agents. - GPU-rich/GPU-poor division of labor: an agent was regularly compute-starved and switched to writing specs, byte-math, and acceptance analysis for other GPU-rich agents to execute. Some agents offered external Modal compute for another agent blocked DFlash training. - Cross-agent kernel debugging: an agent debugged another agent run of of yet another agent fused drafter: found a Triton store/load aliasing race in _k_qnorm_rope, a second shape bug, then rewrote attention with flash-decoding split-KV. Fixes posted "take freely." - Quota-pooling norm: Often agents would stage a candidate publicly for whoever has quota to run it. Agents will then usually credits the originator. This behavior emerged because of the 10-job/24h cap (e.g. pupa's package run by resystagent and fabulous-frenzy). Discoveries & reversals: - Agents would make many discoveries and reversal of them, giving them names like the following: - 127 TPS "wall" was an artifact. a mathematical proof of the max possible speed became called in the community the "int4-Marlin floor" but a later agent called the proof circular (only varied the bandwidth term, never overhead). Finally another agent broke to 247 TPS via MTP speculative decoding on a vLLM nightly. - "Smarter draft loses." An agent showed that a 2B drafter's ~1 GB/token read dominates even at perfect acceptance and a much smaller 256-hidden drafter wins at batch-1 because its weights are nearly free to read. Agent discussed how per-accepted-token cost ≈ draft bytes read / acceptance. - "DFlash near-random acceptance": an agent remotly diagnosed the 2–5% acceptance rate of another agent as near-random, ruling out undertraining/vocab caps and pointing to a train/serve hidden-state mismatch (bf16 E4B extraction vs int4 serving). - Much of the race was noise: one agent decide to run the #1# submission 4 times and found a σ≈1.16 TPS variation in single run. Another agent confirmed across 358 runs / 66 buckets: frontier deltas <~4 TPS are ties. Community adopted a significance norm. So many interesting interactions in the interaction board: You can explore also the lineage of inventions from the agents at: And the challenge it-self at And the organization behind the challenge at

bitcoin:native 现货 ETF 6月18日 总持仓降至 1,247,352.18 BTC，当日净流出 1,784.05 BTC，主要流出来自贝莱德 -1,540.82 BTC，唯一净流入来自摩根士丹利 +166.24 BTC。 本周截至 6月18日累计净流出 3,476.96 BTC，较上周 6月8日至6月12日净流出 6,183.81 BTC 明显收窄，资金压力减少约 2,706.84 BTC。 #Bitget# 来了就是VIP！Crypto、美股、CFD，全球先机一站布局

泄露韩国 64% 的公民购物数据后，韩国电商网站酷澎 (#Coupang#) 被罚款 6,247 亿韩元，约合人民币 27.7 亿元。 酷澎泄露的数据涵盖 3,370 万名客户，泄露源头则是已经离职 1 年多的前工程师，这名工程师本身也不是韩国人，离职后使用窃取的加密签名密钥持续访问服务器获得数据。 查看详情：

GitHub 爆火项目榜：本周新增星标精选 TOP 7。 我最喜欢推荐7 screenpipe 最符合我对 AI 时代个人工具的期待，它不是一次性玩具，而是能持续积累价值的“第二大脑”。 这些项目在本周星标增长明显，涵盖 AI 视频生成、虚拟伴侣、安防监控、物理仿真、照片管理等领域。 1 MoneyPrinterTurbo +7948 【造富潜力】 仓库： 简介：利用 AI 大模型，一键生成高清短视频。输入主题或关键词即可自动完成文案、素材、字幕、配乐并合成完整视频，支持中英文。 亮点：全自动流水线 + WebUI，适合内容创作者快速生产。 造富潜力：项目名称本身带有“印钞机”意味，社区常将其视为批量生成短视频的工具，适合尝试制作无脸视频或带货内容并上传抖音、YouTube、TikTok 等平台进行流量变现。但实际效果高度依赖视频质量、平台算法、内容合规以及个人运营能力。目前社区讨论多为潜力宣传，尚未有大规模公开验证的成功收益案例。 2 airi +1079 仓库： 简介：自托管的 AI 虚拟伴侣，支持实时语音聊天、可玩游戏（Minecraft / Factorio 等），兼容 Web / macOS / Windows。 亮点：完全本地部署，可自行控制。 3 Frigate +1308 仓库： 简介：实时对象检测的本地 NVR，专为 IP 摄像头设计，集成 Home Assistant，使用本地 AI 完成物体识别。 亮点：隐私优先，无云端依赖，智能家居用户常用。 4 genesis-world +247 仓库： 简介：通用机器人与具身 AI 的仿真平台，集成多物理引擎和照片级渲染器，用于物理 AI 开发。 亮点：生成式世界模拟，适合机器人和强化学习研究。 5 ppf-contact-solver +1603 仓库： 简介：物理仿真接触求解器，支持布料、实体和杆的复杂物理模拟。 亮点：高性能接触求解，适用于游戏开发、材料仿真等领域。 6 PhotoPrism +68 仓库： 简介：AI 驱动的自托管照片管理应用，类似私有版 Google Photos，支持自动标签、智能搜索和 RAW 处理。 亮点：完全本地运行，隐私优先。 7 screenpipe +146 仓库： 简介：本地屏幕 + 音频持续录制工具，24/7 捕获数字生活，生成可搜索的 AI 记忆库。 亮点：把电脑变成有记忆的个人 AI 引擎，支持后续个性化工作流。

【M-line Music】 【M-line Music#247】M-line# Sp ～LOVELY～「バカにしないで」／稲場愛香「LOVE is BLIND」／キッチン／エムハロ vol.9 ツアー日記 MC 小関舞・村越彩菜 #mlinemusic#

【M-line Music#247】M-line# Sp ～LOVELY～「バカにしないで」／稲場愛香「LOVE is BLIND」／キッチン／エムハロ vol.9 ツアー日記 MC 小関舞・村越彩菜 #譜久村聖# #竹内朱莉# #森戸知沙希# #小関舞# #稲場愛香# #高瀬くるみ# #村越彩菜# #ハロプロ# #mlinemusic#