Intel 情報源

Physics-Informed Deep Learning for False Ventricular Tachycardia Alarm Reduction in the ICU

物理資訊深度學習減少ICU假警報。

Y Combinator Early Access Network

Y Combinator推出早期訪問網絡。

Transformers as In-Context Samplers: From Closed-Form Diffusion to Estimation-Free Sampling

Transformer可作為資料的上下文取樣器。

Omni Interaction Agent Technical Report

Gander統一全感知、即時互動等。

Good Pretraining, Bad SFT: Checkpoint Quality Across the Training Stack

最佳預訓練檢查點在完整SFT後未必最佳。

PlannerForge: LLM Agents for Scenario-Based Testing of Motion Planners in Autonomous Driving

PlannerForge使用LLM代理進行基於場景的ADS測試。

Do you think it happened? Research stolen from their Codex private chats

指控從私人聊天中竊取研究。

SQLMorph: Query Mutation and Fine-Grained Metrics for Text-to-SQL Evaluation

SQLMorph變異查詢並提供細粒度指標…

LG TVs caught spying even when offline or on standby

LG電視據報即使關機或閒置也會監視。

ONE CYLinder: A Benchmark for Graph-Based Surrogate Modeling of Unsteady Bluff-Body Flows

用於圖形代理建模的ONECYL基準…

SkillAdam: Stable and Efficient Skill Evolution for Agents

SkillAdam以Adam啟發的方式優化代理技能…

Evaluating and Improving Evidence-Grounded Fact-Checking in LLMs via Multi-Round Evidence Ablation

LLM在…中依賴參數知識而非證據。

AuK Technical Report: An Open-Source Foundational Model for Speech Generation and Editing

AuK是語音生成和…的開源模型。

Open Source Durable Objects for Postgres

When Models Defer to Wrong Answers: A Robustness Audit of Source-Attributed Cues in Multiple-Choice QA

來源歸因提示可使LLM服從錯誤答案。

C*: Unifying Programming and Verification in C

C*語言結合程式設計與形式驗證。

Into the depths of C: Elaborating the de facto standards (2016)

探討2016年起的實際C標準。

Experience Funnel: A State-Policy Alternating Loop for Self-Evolving Agents

自演化代理交替快速狀態和慢速策略…

US police fear Meta smart glasses could be used to secretly record them

警方擔憂Meta眼鏡可秘密錄影。

The BatchNorm Illusion: Diagnosing Normalization Artifacts in Machine Unlearning Evaluation

BatchNorm偽影混淆遺忘評估…

From Scores to Evidence: Auditable Decisions Can Improve Speech Deepfake Detection

可審計決策記錄改善語音深偽…

Evolution of Multimodal Question Answering: From Modality-Adaptive Extraction to Unified Language Representation

比較多模態QA框架，顯示轉向統一…

Q2D-Web: A Large-Scale Benchmark for Retrieval in Agentic RAG Systems

大型基準測試用於代理式RAG檢索，含1.9億…

I Rewrote My Back End in Go in 72 Hours and Cut My AWS Bill by 94%

用Go重寫後端，72小時內AWS成本降94%。

Disappointed Optimists

反思未達期望與樂觀。

OntoKG-EQ: A provenance-grounded, competency-question-governed knowledge graph for auditable analyst querying

知識圖譜用於新興領域中可稽核的分析師查詢…

API Benchmark Scores Do Not Reliably Transfer to Chatbot Interfaces

API基準分數無法可靠轉移至聊天機器人…

Reverse engineering the 2008 LogiCola, a logic drill program

逆向工程2008年邏輯鑽孔程式。

Google DeepMind Releases AlphaGenome Atlas

DeepMind發布AlphaGenome Atlas用於基因組學。

ZX Spectrum: Experimenting with 1-Bit Sound

ZX Spectrum上1位元聲音實驗。

Closing the Consistency Gap: Self-Evolving Agents That Learn to Stay on Course

自演化代理縮小任務成功的一致性差距。

Benchmarking Qwen3.8 27B quantizations: 4-bit holds up, 1-bit collapses

Qwen3.8 27B：4位元可用，1位元失敗。

A Closed-Form Estimator and Diagnostic Battery for Anchor-Judge Error Correlation, Under a Single-Common-Factor Model

錨定評審誤差相關性的閉式估計器。

The two Christian saints who are the Buddha

兩位基督教聖人與佛陀起源相關。

DHS 'Predictive Policing' Unit Is Analyzing Americans' Financial Habits

DHS分析財務習慣用於預測性警務。

Paramount Caught Using 'Astroturf' Group to Drum Up Fake Support for Merger

Paramount使用假團體支持合併。

The far right's win in Germany is sending a chill through Europe's establishment

德國極右翼勝選令歐洲建制派擔憂。

Apple unfolds its decade-long iPhone

蘋果公布十年iPhone計畫；澳洲推動…

Ostrich: Taking Large Strides Through Stiff Contact in Differentiable Dynamics

可微分模擬器處理大規模剛性接觸…

The Rater Ising-Potts Model with LLM-Derived Weights: An Application to Multi-Category Scoring Reliability

Ising-Potts模型用於多類別評分可靠性。

Hi-FLoop: Hierarchical State-Feedback Loops for Multi-Timescale World Modeling

多時間尺度世界模型的分層狀態回饋迴路。

Evidence-Grounded Retrieval for Investigation Hunt Lead Generation from CTI Reports

基於證據的檢索用於CTI報告的獵捕線索。

Silent Revision: Measuring Undisclosed Change in the Safety Frameworks of Frontier AI Developers

衡量AI安全框架中未揭露的變更。

Latest open artifacts (#24): Motif-3, GLM-5.3, Hy4-preview and open model licenses

新開放模型與授權擴展生態系。

AlphaGenome Atlas predictive map of every DNA letter change in the human genome

地圖預測所有人類DNA變化的影響。

I-have-ADHD: A skill to stop coding agents from burying the answer

防止編碼代理隱藏答案的技巧。

It's All in the Way You Say It: The Role of Information Representation in LLM-Based Glycemic-Event Prediction

資訊表示影響LLM血糖事件…

LibreOffice breaks download records after declaring it has no AI features

LibreOffice下載創紀錄，無AI功能。

Benchmark Scores Are Pipeline-Dependent: A Reliability Audit of Cybersecurity LLM Benchmarks

網路安全LLM基準分數依賴管線。

Kairos: A Dataset for Fine-Grained Video-Language Modeling over Space, Time, and Dynamics

用於時間上細粒度影片語言建模的資料集。

We Must Return to the Office to Use AI in Person

主張回辦公室以使用AI的論點。

DaVinci Resolve 21.1

DaVinci Resolve 21.1發布，含更新。

Show HN: Copperhead – Hardware as Fast as Software

Copperhead目標是硬體速度的軟體。

GoAnt: Quality-Diversity Multi-Agent Search for Alpha Factor Discovery in Market Microstructure Data

用於alpha因子的品質多樣性多智能體搜尋…

Chimaera: A Mixture-of-Graph-Experts Architecture for Cross-Task and Cross-Dataset Graph Learning

用於跨任務/資料集的專家混合圖模型…

The first GPT-6 model

首個GPT-6模型及其預期用途。

TontaubeV1: Streaming Text-to-Speech with Hierarchical Codec Modeling and Bounded Context

具有分層編解碼器和有界上下文的串流TTS。

Record Grouping Controls Evidence Weight in Language Models

記錄分組控制語言模型中的證據權重。

Global Divergence, Local Convergence: Representation Geometry in SSMs and Transformers

SSM和Transformer全域不同但局部收斂。

When Victorian Becomes a Prompt: Literary Periodization as a Generative Constraint in 100 AI-Generated Novels

AI生成小說中的維多利亞提示改變文法。