it's crazy to think opus 5.5 itself is the product of an even stronger internal model better than fable 5.1 while being smaller than opus 5.0... whatever anthropic is using internally for training/research must be insanely capable -- and even more insane is that we could have this level of capability in open models running locally within 6-12 months

🧩 Qwen3.8-Flash-Next: A 6B-Active Preview of the Qwen4 Architecture On the same night Zhipu's GLM-5.3-Flash took over the timeline, @Alibaba_Qwen open-sourced Qwen3.8-Flash-Next — explicitly positioned as a preview of the Qwen4 architecture. Zhihu contributor Kitt在进化 argues that for people who actually run models locally, this is the more practical release of the night. The local-deployment groups he is in are, in his words, on fire. His take: the model is called 3.8, but the architecture is really Qwen4 in preview — and it borrows the best ideas from across the field. 1️⃣ Smaller, cheaper, and realistic for local deployment Flash-Next has 125B total parameters with only 6B active — far smaller than the 300B-class GLM-5.3-Flash. The API is priced at ¥1 input, ¥3 output, and ¥0.1 per cached million tokens, roughly matching DeepSeek-V4-Flash's off-peak rates. His comparison, per 1M tokens in RMB — input / output / cache hit: 🔹 Qwen3.8-Flash-Next: 1.0 / 3.0 / 0.1 🔹 GLM-5.3-Flash: 0.4 / 1.4 / 0.115 (limited-time promo rate) 🔹 DeepSeek-V4-Flash (off-peak): 1.5 / 4.5 / 0.05 2️⃣ A Qwen4 preview wearing a Qwen3.8 name The author points out this is the same play Zhipu made: GLM-5.3-Flash's architecture is also completely different from GLM-5.3. Shipping the new architecture as open weights early is deliberate pathfinding. Inference frameworks like vLLM and SGLang, plus the quantization toolchain, all need lead time to adapt before Qwen4 proper arrives. 3️⃣ An architecture that borrows from everyone The author reads Flash-Next as a synthesis of the field's best recent ideas. 🔹 Attention: Qwen's in-house QSA, built to balance throughput and speed on long context. 🔹 Knowledge: it absorbs DeepSeek's Engram line of work, packing prior knowledge into 51B of N-gram side parameters — high knowledge density at minimal compute cost. 🔹 Training: the Muon optimizer, popularized by Kimi, scheduled together with AdamW. 4️⃣ Half the active parameters, still overtaking At 6B active, Flash-Next is nearly half the size of DeepSeek-V4-Flash — yet Qwen's reported benchmarks show it overtaking that model on multiple coding and agent leaderboards. The author says he is not worried about real-world experience: the recent Qwen3.8-27B already proved itself in daily use, and this sits on the same foundation. 🔗 Full Reading: #Qwen# #Alibaba# #Qwen4# #OpenWeights# #LLM# #AIInference# #MoE#

GENERAL MILLS $GIS Q1’27 EARNINGS HIGHLIGHTS 🔹 Revenue: $4.4B (Est. $4.35B) 🟢; -3% YoY 🔹 Adj. EPS: $0.75 (Est. $0.72) 🟢; -13% CC 🔹 Adjusted Oper Profit: $634M (Est. $604M) 🟢 🔹 Organic Net Sales Growth: Flat Reaffirms FY27 Guide: 🔹 Organic Net Sales: -1.5%-+0.5% 🔹 Adjusted Operating Profit: -13% to -8% CC 🔹 Adj. EPS: $3.00-$3.20/share 🔹 FCF Conversion: ~95% of adjusted after-tax earnings Segment Net Revenue: 🔹 North America Retail: $2.5B; -3% organic 🔹 International: $794.3M; +4% organic 🔹 North America Pet: $612.8M; flat organic 🔹 North America Foodservice: $523.1M; +4% organic Other Q1 Metrics: 🔹 Organic Net Sales Growth: Flat 🔹 Adjusted Operating Profit Growth: -11% CC 🔹 Adjusted Gross Margin: 33.3%; -90 bps YoY Comments: 🔸 “we remain confident in our ability to deliver our fiscal 2027 guidance.”

GENERAL MILLS $GIS JUST REPORTED Q1 EARNINGS - Revenue: $4.4B, beating est. of $4.35B 🟢 - Adj EPS: $0.75, beating est. of $0.72 🟢 (-13% CC) - Adjusted Oper Profit: $634M - Organic Net Sales Growth: Flat - North America Retail: $2.5B (-3% organic) - International: $794.3M (+4% organic) - North America Pet: $612.8M (flat organic) - North America Foodservice: $523.1M (+4% organic) Reaffirms FY27 Guidance: - Organic Net Sales: -1.5%-+0.5% - Adjusted Operating Profit: -13% to -8% CC - Adj EPS: $3.00-$3.20/share - FCF Conversion: ~95% of adjusted after-tax earnings

Foreign born population of each country : Jan 2001 Jan 2025 🇦🇹 8.7% 22.5% 🇧🇪 8.4% 20.2% 🇩🇰 4.8% 14.4% 🇫🇷 5.5% 14.0% 🇩🇪 8.9% 20.5% 🇬🇷 6.9% 11.0% 🇮🇸 3.1% 21.8% 🇮🇪 4.0% 23.3% 🇱🇺 37.5% 51.5% 🇳🇱 4.1% 16.8% 🇳🇴 4.1% 18.7% 🇸🇮 2.1% 15.5% 🇪🇸 3.4% 19.3% 🇸🇪 5.3% 20.8% 🇬🇧 4.3% 20.0% Now add children born to foreign parents and the situation looks even worse. We are living through the demographic annihilation of the people of Europe.

The global cryptocurrency market capitalization stands at approximately $2.28–2.37 trillion, having declined slightly by about 0.5%–1.4% over the past 24 hours. Trading volume remains around $4.5–5.5 billion. Bitcoin’s dominance rate is approximately 57.8%–57.9%, while Ethereum’s is about 10.5%. MicroStrategy may pause its Bitcoin accumulation, ending its 13-week streak of consecutive purchases, which has drawn market attention. I do not believe BTC can stabilize at this level; I recommend staying out of the market and waiting on the sidelines.

📌 VI. Trigger Matrix (V2.0 – Observation Status Log) Observation Item Current Value Threshold Status Consecutive Days/Trend Super-Capital Concentration Risk 9.3 8.0 ESCALATION ↑ 1 day (new) AI Governance Risk 8.8 8.0 ESCALATION ↑ 1 day (new) Resilience Ratio 0.63 0.70 ESCALATION ↑ 4 days US-Iran Deal Signing Status 接近 Formal Signing WATCH — Brent Crude Oil Price 3-mo low — THRESHOLD_CROSSED 1 day New Ebola Health Zone (DRC) Confirmed spread — THRESHOLD_CROSSED 3 days EU Accession Talks Launched — THRESHOLD_CROSSED 1 day Items Near Threshold (Elevated Observation): Observation Item Current Value Threshold Current Status • Formal signing of US-Iran deal 接近 Formal Signing ALERT • SpaceX market cap stability Above $2T Drop below $2T WATCH • OpenAI probe scope expands Multi-state Federal involvement ALERT • G7 Summit statements on AI & trade 即将 held Substantive regulatory共识 WATCH • Cross-border Ebola spread Risk rising First邻国 confirmed case ALERT • Clustered cases in fan zones No reports Confirmed cluster transmission WATCH --- 📅 VII. Key Observation List for the Next 72 Hours Grade A Observations (High Impact): Observation Item Potential Impact if Triggered 1. Formal signing of US-Iran MOU Geopolitical entropy pressure declines further, but execution risk仍需 assessed. 2. SpaceX market cap stability above $2T Test of sustainability for super-capital concentration narrative. 3. OpenAI probe expands to federal level Potential further upgrade to AI governance risk level. 4. G7 Summit statements on AI & trade First collective test of institutional response capacity. Grade B Observations (Medium Impact): Observation Item 1. Expansion of Ebola outbreak zone in DRC 2. Subsequent日程 for EU accession negotiations 3. Public health data during FIFA World Cup --- 📜 VIII. CRI Calculation Summary (V1.6) Variable Weight Risk Score Weighted Contribution V_capital 20% 9.3 1.86 V_tech 18% 8.8 1.58 V_inst 18% 8.1 1.46 V_geo 15% 7.5 1.13 V_human 10% 7.6 0.76 V_expansion 8% 7.9 0.63 V_market 6% 7.2 0.43 V_energy_price 5% 6.5 0.33 Total 100% CRI = 8.2 Calibration Notes: Added V_capital variable (weight 20%) to reflect super-capital concentration as a new structural risk dimension. V_tech上调 to 8.8 (AI governance race launch). V_geo下调 to 7.5 (US-Iran deal接近, declining war risk). --- 📌 IX. Structural Conclusion On June 13, 2026, the global civilizational system's Resilience Ratio remains below the 0.70警戒线 for the fourth consecutive day. What is most worth recording today is not war – but the first time in human civilization that private wealth approaches the GDP of a中等发达国家. When a single entrepreneur owns a satellite network, rocket system, AI platform, energy network, financial capital, and global data流入口, civilization is entering a new organizational form: Transitioning from nation-state-led civilization to platform-infrastructure-led civilization. If the core question of the 20th century was "How to constrain state power?", then the core question of the latter half of the 21st century may well become "How to govern super-platform power."

ช่วงเวลาบูธกิจกรรมนะคะ♪ ⏰เวลา15.30~19.00 💜ྀིRina Merchandise🩷ྀི 👑Special Photo Set👑 (ในเซ็ตมีรูป5ใบ) (ขาย Type ละ 20 เซ็ตเท่านั้น) 450฿ +250฿ ลายเซ็น+เขียนข้อความ • Type A Thai Dress 👗🩵 • Type B White Onepiece 👗🤍 #rinaIzuta# #伊豆田莉奈# #いずりな#

SWE-Together Update: Claude Fable 5 and 5.1 are the new top models on SWE-Together, and Meta's Muse Spark 1.3 is the best value on the board. SWE-Together is our benchmark of 109 real coding tasks with a simulated user in the loop, built to tell you whether a new model will actually match your expectations before you hand it your work. Six quick findings from the updated board: 1. Fable 5 beats 5.1 on consistency, but 5.1 is more efficient. At their best the two actually perform the same. The gap is in the worst runs, where Fable 5.1 scored zero on 14 trials and Fable 5 on only 6. However, 5.1 is 25% faster, about 2x cheaper, and needs slightly fewer corrections from the user (1.45 vs 1.53 corrective messages per task) than Fable 5. 2. Muse Spark 1.3 is the value outlier. Per solved task, it is about 2.5x cheaper than Fable 5.1 and 5x cheaper than Fable 5. Performance-wise, it ties with Fable 5 and 5.1 for best on Python (73%) and Rust (100%). 3. Newer is not automatically better at collaborative coding. GPT-5.6-Sol is the newer model, yet it needs more corrections from the user than GPT-5.5 (1.66 vs 1.59 per task). 4. "Stronger models need less steering" is a trend, but not guaranteed. With more models added, the correlation between pass@1 and user correction weakens from -0.92 to -0.72, and Fable 5 takes more corrections than Opus 4.8 (1.53 vs 1.38) despite a 6 pp higher pass@1. 5. Frontier progress contributes greatly to stability. Fable 5 converts 89% of its pass@1 into pass² (both runs solve), versus 60 to 67% for DeepSeek V4 Pro, GLM-5.1, and MiniMax 2.7. 6. There is still plenty of headroom. 16 of 109 tasks have zero solves across all tested models, and even Fable 5 at 69% pass@1 sits about 9 pp below the ~78% that the original human patches scored. The full leaderboard, with per-task and per-trial breakdowns, is at and the benchmark is open source at