#Codebook
The Aid Worker Security Database (AWSD) codebook is now available in English, French, Spanish and Arabic here - humanitarianoutcomes.org/AWSD-codebook

@globalaffairs.canada.ca @minbz.bsky.social @unocha.org

#WHD2026 #ActForHumanity
September 30, 2026 at 4:27 PM
Dae Ung Jo, Jongin Lim, YoungJoon Yoo, Daeho Um: Codebook-Guided Cross-Modal Knowledge Distillation for Structurally Heterogeneous Features https://arxiv.org/abs/2609.37243 https://arxiv.org/pdf/2609.37243 https://arxiv.org/html/2609.37243
September 30, 2026 at 6:42 AM
Wasif Ali, Yen-Ming Chen, Fan-Shuo Tseng: DNN-Assisted Near-Optimal Codebook Design for Uplink SCMA Schemes over Rician Fading Channels https://arxiv.org/abs/2609.36137 https://arxiv.org/pdf/2609.36137 https://arxiv.org/html/2609.36137
September 30, 2026 at 6:42 AM
新たなSpectre v2攻撃「BTR」により、Intel・AMD・Arm製CPUがデータ漏洩リスクに晒される恐れ | Codebook|Security News https://codebook.machinarecord.com/threatreport/silobreaker-cyber-alert/47855/
September 30, 2026 at 5:28 AM
テック企業の機微データ写したスクリーンショットを、AIモデルが公開リポジトリに投稿 | Codebook|Security News https://codebook.machinarecord.com/threatreport/silobreaker-cyber-alert/47858/
September 30, 2026 at 5:27 AM
FineSID: Scalable and Efficient Semantic Identifier Learning for Generative Recommendation

Softens hard Top-1 codeword assignment in semantic ID learning so gradients reach the whole codebook, easing codebook collapse while keeping IDs discrete.

📝 arxiv.org/abs/2609.36670
FineSID: Scalable and Efficient Semantic Identifier Learning for Generative Recommendation
A critical prerequisite of generative recommendation is designing semantic identifiers (SIDs) that are both scalable to large item sets and efficiently learnable. Existing SID learning methods fundame...
arxiv.org
September 30, 2026 at 4:12 AM
🎧 Code audio and video directly, picture and sound together, or transcribe first with the new qlm_transcribe() function!

✅ Every response is validated against your codebook. One that doesn't fit is a recorded failure, not a quiet NA

🌐 Any OpenAI-compatible model, plus web-search tools
September 30, 2026 at 1:11 AM
Plan next school year before it plans you.

Set up your 2026-27 planner in one sitting. Save this for back-to-school season. #planner #backtoschool #studygram #organization #studentlife

🛒 Instant printable — grab it 👇

https://revenue-cockpit.netlify.app/store/journeyman-electrician-nec-codebook…
September 29, 2026 at 6:45 AM
Codebook of Mormon is this anything
September 29, 2026 at 2:00 AM
The Youth Risk Behavior Survey 2025 is out.
- Q63 (transgender identity) & Q94 (ask consent) were deleted from data
- Codebook omits the usual appendix B showing survey-weighted and raw frequencies of all answers for all questions.
- Usually provides SAS, SPSS, and Access, but no SPSS this year.
September 28, 2026 at 6:35 PM
New on CRAN: mariposa (0.7.3). View at https://CRAN.R-project.org/package=mariposa
mariposa: 'SPSS'-Compatible Statistical Tools for Survey Data
Statistical analysis of survey data with full support for survey weights, grouped operations, and 'tidyverse' integration. Provides 80 functions for data import/export ('SPSS', 'Stata', 'SAS', 'Excel') with label roundtripping and tagged NA preservation, label management (variable labels, value labels, type conversions, missing value declaration), data transformation (recoding, dummy coding, standardization, centering), descriptive statistics, codebook generation, hypothesis testing, correlation analysis, post-hoc comparisons, weighted statistics, scale analysis, regression, non-parametric tests, exact tests, factorial ANOVA, and ANCOVA. Every analysis offers compact print() and detailed summary() output with toggleable sections. Statistical results are validated against 'SPSS' version 29 within documented per-tier tolerances (see the compatibility vignette for per-function status). Methods follow the published algorithms of IBM Corp. (2023, "IBM SPSS Statistics Algorithms"), the Lilliefors-corrected normality test of Dallal and Wilkinson (1986) &lt;<a href="https://doi.org/10.1080%2F00031305.1986.10475419" target="_top">doi:10.1080/00031305.1986.10475419</a>&gt;, and the adjusted standardized residuals of Haberman (1973) &lt;<a href="https://doi.org/10.2307%2F2529686" target="_top">doi:10.2307/2529686</a>&gt;. Designed for survey researchers, social scientists, and students working with complex survey designs.
CRAN.R-project.org
September 28, 2026 at 5:24 PM
動画圧縮の概念を根本から覆す最新の研究論文が公開されました。

**【論文紹介】GVCC: Zero-Shot Video Compression via Codebook-Driven Stochastic Rectified Flow**

従来の圧縮技術は「元の映像への回帰」を目指すため、超低ビットレート環境では映像がぼやけてしまう課題がありました。

本研究が提案する**GVCC**は、生成AIモデル(Rectified Flow)をデコーダーとして直接活用するゼロショット圧縮フレームワークです。

**革新的なアプローチ:**
* **「回帰」から「生成...
September 28, 2026 at 9:34 AM
New CRAN package mariposa with initial version 0.7.3
#rstats
https://cran.r-project.org/package=mariposa
CRAN: Package mariposa
Statistical analysis of survey data with full support for survey weights, grouped operations, and 'tidyverse' integration. Provides 80 functions for data import/export ('SPSS', 'Stata', 'SAS', 'Excel') with label roundtripping and tagged NA preservation, label management (variable labels, value labels, type conversions, missing value declaration), data transformation (recoding, dummy coding, standardization, centering), descriptive statistics, codebook generation, hypothesis testing, correlation analysis, post-hoc comparisons, weighted statistics, scale analysis, regression, non-parametric tests, exact tests, factorial ANOVA, and ANCOVA. Every analysis offers compact print() and detailed summary() output with toggleable sections. Statistical results are validated against 'SPSS' version 29 within documented per-tier tolerances (see the compatibility vignette for per-function status). Methods follow the published algorithms of IBM Corp. (2023, "IBM SPSS Statistics Algorithms"), the Lilliefors-corrected normality test of Dallal and Wilkinson (1986) &lt;<a href="https://doi.org/10.1080%2F00031305.1986.10475419" target="_top">doi:10.1080/00031305.1986.10475419</a>&gt;, and the adjusted standardized residuals of Haberman (1973) &lt;<a href="https://doi.org/10.2307%2F2529686" target="_top">doi:10.2307/2529686</a>&gt;. Designed for survey researchers, social scientists, and students working with complex survey designs.
cran.r-project.org
September 28, 2026 at 9:02 AM
OpenAIのエージェントがユーザーのアップした画像53件をサードパーティWebサイトへ投稿 | Codebook|Security News https://codebook.machinarecord.com/threatreport/silobreaker-cyber-alert/47844/
September 28, 2026 at 7:57 AM
NetScalerのRCEゼロデイ2件、攻撃での悪用をCitrixが確認(CVE-2026-88771、CVE-2026-88772) | Codebook|Security News https://codebook.machinarecord.com/threatreport/silobreaker-cyber-alert/47848/
September 28, 2026 at 7:57 AM
OpenAIのエージェントがオーストラリア政府サイトをハッキング | Codebook|Security News

https://www.walknews.com/1389995/

OpenAIのエージェントがオーストラリア政府サイトをハッキング Help Net Security – September 24, 2026 OpenAIのエージェントは、Hugging FaceやRubyGemsのハ [...]
OpenAIのエージェントがオーストラリア政府サイトをハッキング | Codebook|Security News - WALK NEWS
OpenAIのエージェントがオーストラリア政府サイトをハッキング
www.walknews.com
September 26, 2026 at 8:00 AM
El modo de cifrado GCM se ha consolidado como estándar en HTTPS por su eficacia y seguridad

🤖 IA: No es clickbait ✅
👥 Usuarios: No es clickbait ✅

#criptografía #https #gcm

👇👇👇
El modo de cifrado GCM se ha consolidado como estándar en HTTPS por su eficacia y seguridad
La noticia explora la evolución de los modos de cifrado por bloques, centrándose especialmente en cómo GCM (Galois/Counter Mode) se ha convertido en el estándar dominante para HTTPS. El artículo comienza explicando que el cifrado por bloques, como AES, opera con bloques de tamaño fijo (128, 192 o 256 bits), pero para cifrar datos más largos se requieren modos de operación adicionales. Se analizan varios modos históricos: ECB (Electronic Codebook), considerado inseguro porque cifra bloques de forma independiente sin aleatoriedad, lo que permite identificar patrones en datos estructurados como imágenes; y CBC (Cipher Block Chaining), que, aunque mejor que ECB, presenta vulnerabilidades como los ataques de padding oracle (Lucky13 en 2013 y POODLE en 2014), lo que llevó a su eliminación en TLS 1.3. El artículo destaca que CTR (Counter Mode) es un modo más seguro que CBC, pero carece de autenticación integrada. GCM, una evolución de CTR, añade autenticación mediante GMAC (Galois Message Authentication Code), garantizando integridad y autenticidad del mensaje. Esto lo hace especialmente adecuado para HTTPS, VPNs y comunicaciones seguras modernas. Además, se menciona que ChaCha20-Poly1305 es una alternativa que se está adoptando en dispositivos móviles por su eficiencia en software sin aceleración hardware. El artículo concluye que AES-GCM es el modo de cifrado dominante en HTTPS, especialmente en entornos con soporte para aceleración por hardware, pero reconoce que la tecnología sigue evolucionando. La noticia es claramente técnica, dirigida a profesionales de la seguridad informática y desarrolladores, y proporciona una explicación clara de los conceptos clave sin recurrir a sensacionalismo.
es.killbait.com
September 26, 2026 at 6:17 AM
OpenAIのエージェントがオーストラリア政府サイトをハッキング | Codebook|Security News

https://www.wacoca.com/news/2941647/

OpenAIのエージェントがオーストラリア政府サイトをハッキング Help Net Security – September 24, 2026 OpenAIのエージェントは、Hugging FaceやRubyGemsのハ [...]
OpenAIのエージェントがオーストラリア政府サイトをハッキング | Codebook|Security News - News | wacoca.com
OpenAIのエージェントが5〜6月にオーストラリア政府の公衆衛生Webサイトを含む3つのサイトをハッキングしようとしていたと独立系リサーチラボTransluceが報告。オーストラリア首相も、政府機関が運営するメディケア統計報告ポータルが同社エージェントに侵害されて公開・非公開ファイルがアクセスを受けたと伝えた。
www.wacoca.com
September 26, 2026 at 12:00 AM
OpenAIのエージェントがオーストラリア政府サイトをハッキング | Codebook|Security News https://codebook.machinarecord.com/threatreport/silobreaker-cyber-alert/47831/
September 25, 2026 at 8:05 AM
ShinyHuntersがOracle PeopleSoftのゼロデイを悪用してFBIからデータ盗んだと主張 | Codebook|Security News https://codebook.machinarecord.com/threatreport/silobreaker-cyber-alert/47823/
September 25, 2026 at 8:05 AM
ShinyHuntersがClopのリークサイトをハッキング&改ざん | Codebook|Security News https://codebook.machinarecord.com/threatreport/silobreaker-cyber-alert/47819/
September 25, 2026 at 8:05 AM
AIエージェントを使用した攻撃でクレジットカード情報60万件分が盗まれる | Codebook|Security News https://codebook.machinarecord.com/threatreport/silobreaker-cyber-alert/47815/
September 25, 2026 at 8:05 AM
Roundcubeの重大な脆弱性、攻撃で悪用されるように:CVE-2026-48842 | Codebook|Security News https://codebook.machinarecord.com/threatreport/silobreaker-cyber-alert/47835/
September 25, 2026 at 8:04 AM
WordPressの重大な脆弱性、公開直後に悪用される(CVE-2026-87902) | Codebook|Security News https://codebook.machinarecord.com/threatreport/silobreaker-cyber-alert/47839/
September 25, 2026 at 8:04 AM
A survey scale gets alpha, item-total correlations, a line in the methods. The codebook behind a qualitative study does the same job, turning what people said into a claim, and usually gets no test at all.
https://relicheck.com/signal/your-codebook-is-an-instrument-too.html
Your Codebook Is an Instrument Too · The Signal
Your Codebook Is an Instrument Too
relicheck.com
September 24, 2026 at 11:04 PM