「bias」の検索結果

Detects allele-specific transcription factor or histone modification binding from heterozygous-variant ChIP-seq using WASP (reference-bias filter; mandatory upstream), RASQUAL (joint QTL + bias-corrected testing), BaalChIP (Bayesian beta-binomial with copy-number-aware overdispersion), and AlleleSeq (personalized diploid genome). Handles imprinted-locus awareness, X-inactivation artifacts, cancer copy-number imbalance, and integration with downstream caQTL / bQTL mapping. Use when identifying variants with allelic effects on TF binding, fine-mapping causal regulatory variants, validating deep-learning variant predictions, or characterizing cis-acting regulatory effects.

日本語の概要は準備中です。原文の説明を表示しています。

Advanced Evaluation workflow skill. Use this skill when the user needs This skill should be used when the user asks to "implement LLM-as-judge", "compare model outputs", "create evaluation rubrics", "mitigate evaluation bias", or mentions direct scoring, pairwise comparison, position bias, evaluation pipelines, or automated quality assessment and the operator should preserve the upstream workflow, copied support files, and provenance before merging or handing off.

日本語の概要は準備中です。原文の説明を表示しています。

Advanced Evaluation workflow skill. Use this skill when the user needs This skill should be used when the user asks to "implement LLM-as-judge", "compare model outputs", "create evaluation rubrics", "mitigate evaluation bias", or mentions direct scoring, pairwise comparison, position bias, evaluation pipelines, or automated quality assessment and the operator should preserve the upstream workflow, copied support files, and provenance before merging or handing off.

日本語の概要は準備中です。原文の説明を表示しています。

管理咨询 (管理咨询 (Management Consulting) — 战略与管理咨询的职业认知操作系统，从业者/咨询顾问/想入行者/采购咨询服务的客户视角。覆盖: (a) 第一性张力 — 假设驱动 (hypothesis-driven / answer-first / Day-1 answer) 的「先有答案再验证」⇄ 数据驱动 (data-led / bottom-up) 的「先穷尽事实再归纳」，这行的核心张力是「先射箭再画靶 vs 先画靶再射箭」，资深人偏 top-down 假设树减少 boil-the-ocean，但要时刻防 confirmation bias; 以及更深层的张力 — 「卖洞见 (insight / so-what) vs 卖工时 (analysis / 体力)」，「客户影响 (client impact / 落地) vs 智识严谨 (intellectual rigor / 漂亮 deck)」; (b) 方法论正典 (最标准化、最易蒸出高质量) — MECE (相互独立完全穷尽)、issue tree / logic tree (问题树/逻辑树)、hypothesis-driven problem solving (假设驱动)、Pyramid Principle (Minto 金字塔原理: SCQA 情境-冲突-疑问-回答 + 结论先行 + 自上而下 + 归纳/演绎)、80/20 (帕累托/抓大放小)、so-what (所以呢/洞见提炼)、storyline (故事线/ghost deck 鬼影稿)、driver tree / profit tree、2x2 矩阵、value chain、frameworks (Porter 五力/3C/4P/7S/BCG 矩阵/价值链) 既是脚手架也是陷阱 (套框架 vs 真洞见); (c) 行业结构 — MBB (McKinsey/BCG/Bain 战略三巨头) vs Big Four 咨询臂 (Deloitte/PwC Strategy&/EY-Parthenon/KPMG) vs 精品战略所 (Oliver Wyman/Kearney/Roland Berger/L.E.K./Arthur D. Little) vs 专业/职能 boutique (运营/数字化/PE 尽调); up-or-out (非升即走)、leverage model (杠杆模型: 1 个合伙人带多个 EM/咨询师/分析师的金字塔)、partner/principal/EM(项目经理)/associate(咨询师)/analyst(分析师) 职级; (d) 项目交付模型 (the case/study/engagement) — 提案/SOW (scope of work)、problem definition (问题定义/真问题 vs 表象问题)、workplan (工作计划)、hypothesis tree (假设树)、analysis (分析/建模)、synthesis (综合)、recommendation (建议)、implementation (落地); 客户管理 (steering committee 指导委员会、stakeholder 利益相关方、对齐); (e) 产出物 — the deck (PPT 幻灯片是咨询的母语)、executive summary、action title (行动标题/每页一句话结论)、ghost deck (先搭骨架后填数据)、Excel 财务模型、framework 图; (f) 职业/招聘 — case interview 生态 (案例面试: Victor Cheng LOMS/Case in Point/market sizing 市场估算/估算题/fit 行为面)、网申-笔试-案例面、up-or-out、exit options (退出路径: PE/VC/corp dev 企业发展/战略岗/创业/industry); (g) 争议/批判 — 咨询的价值之辩 (「借你的表告诉你时间」「frameworks 是剧场表演 vs 真洞见」Matthew Stewart《The Management Myth》)、McKinsey 的丑闻 (Purdue 阿片类/独裁政府/ICE/利益冲突, Bogdanich & Forsythe《When McKinsey Comes to Town》)、咨询 vs 自建战略部、generalist 通才 vs specialist 专才之争、AI/GenAI 对初级分析师工作的冲击; (h) 流派 — 战略经典 (Porter 定位学派/五力 vs 资源基础观 RBV vs 蓝海战略 Kim&Mauborgne vs 颠覆式创新 Christensen vs Mintzberg 应急战略批判 planning school); 三巨头气质分野 (McKinsey 事实驱动/CEO 信任/通才, BCG 概念/思想领导力/经验曲线/增长-份额矩阵, Bain 结果导向/NPS 净推荐值/PE 关系/与客户深度绑定)。不含: 投行 M&A/财务顾问 (估值/承销超出范畴)、纯 IT 实施/系统集成 (SI 人力外包)、纯审计/税务、纯人力 RPO/猎头、内部战略部 only (虽相关但本 skill 聚焦专业服务咨询的外部视角)。) Master OS — automated mastery of 管理咨询 (Management Consulting) — 战略与管理咨询的职业认知操作系统，从业者/咨询顾问/想入行者/采购咨询服务的客户视角。覆盖: (a) 第一性张力 — 假设驱动 (hypothesis-driven / answer-first / Day-1 answer) 的「先有答案再验证」⇄ 数据驱动 (data-led / bottom-up) 的「先穷尽事实再归纳」，这行的核心张力是「先射箭再画靶 vs 先画靶再射箭」，资深人偏 top-down 假设树减少 boil-the-ocean，但要时刻防 confirmation bias; 以及更深层的张力 — 「卖洞见 (insight / so-what) vs 卖工时 (analysis / 体力)」，「客户影响 (client impact / 落地) vs 智识严谨 (intellectual rigor / 漂亮 deck)」; (b) 方法论正典 (最标准化、最易蒸出高质量) — MECE (相互独立完全穷尽)、issue tree / logic tree (问题树/逻辑树)、hypothesis-driven problem solving (假设驱动)、Pyramid Principle (Minto 金字塔原理: SCQA 情境-冲突-疑问-回答 + 结论先行 + 自上而下 + 归纳/演绎)、80/20 (帕累托/抓大放小)、so-what (所以呢/洞见提炼)、storyline (故事线/ghost deck 鬼影稿)、driver tree / profit tree、2x2 矩阵、value chain、frameworks (Porter 五力/3C/4P/7S/BCG 矩阵/价值链) 既是脚手架也是陷阱 (套框架 vs 真洞见); (c) 行业结构 — MBB (McKinsey/BCG/Bain 战略三巨头) vs Big Four 咨询臂 (Deloitte/PwC Strategy&/EY-Parthenon/KPMG) vs 精品战略所 (Oliver Wyman/Kearney/Roland Berger/L.E.K./Arthur D. Little) vs 专业/职能 boutique (运营/数字化/PE 尽调); up-or-out (非升即走)、leverage model (杠杆模型: 1 个合伙人带多个 EM/咨询师/分析师的金字塔)、partner/principal/EM(项目经理)/associate(咨询师)/analyst(分析师) 职级; (d) 项目交付模型 (the case/study/engagement) — 提案/SOW (scope of work)、problem definition (问题定义/真问题 vs 表象问题)、workplan (工作计划)、hypothesis tree (假设树)、analysis (分析/建模)、synthesis (综合)、recommendation (建议)、implementation (落地); 客户管理 (steering committee 指导委员会、stakeholder 利益相关方、对齐); (e) 产出物 — the deck (PPT 幻灯片是咨询的母语)、executive summary、action title (行动标题/每页一句话结论)、ghost deck (先搭骨架后填数据)、Excel 财务模型、framework 图; (f) 职业/招聘 — case interview 生态 (案例面试: Victor Cheng LOMS/Case in Point/market sizing 市场估算/估算题/fit 行为面)、网申-笔试-案例面、up-or-out、exit options (退出路径: PE/VC/corp dev 企业发展/战略岗/创业/industry); (g) 争议/批判 — 咨询的

日本語の概要は準備中です。原文の説明を表示しています。

CEO (Chief Executive Officer) — 一家公司「最高操作员」的工艺, 与泛泛「领导力 / 管理学」不同: (a) 设定战略与凝结愿景 (Drucker 的「企业的经营理论」+ Porter 论定位 + Hamilton Helmer 的「七力」+ Andy Grove 的「战略转折点」+ Roger Martin 《Playing to Win》); (b) 资本配置是 CEO 单一最高杠杆的工作 (Buffett/Munger Berkshire 致股东信 + William Thorndike 《The Outsiders 商界局外人》论 8 位非常规 CEO + Michael Mauboussin 论 hurdle rate + Bezos 年度致股东信「invariants vs bets」+ Mark Leonard Constellation Software); (c) 组建与领导高管班子 (Patrick Lencioni 《团队的五大障碍》+ Ben Horowitz 《艰难的事》+ Reed Hastings 《No Rules Rules》人才密度 + Andy Grove HOP《高产出管理》+ Lou Gerstner IBM 转型 + Marshall Goldsmith); (d) 董事会与投资人关系 (Larcker & Tayan 《公司治理重要》+ Bill George 《True North》+ 正式 CEO-董事长 / CEO-独立董事 protocol + 维权投资人 playbook + 资本市场沟通); (e) 企业文化与价值观作为操作系统 (Netflix 文化手册 + Amazon Leadership Principles + Bridgewater 《Principles》Ray Dalio + Schein 论组织文化 + 任正非 灰度 + 张瑞敏 人单合一 + 稻盛和夫 阿米巴); (f) 高不确定下的高风险决策 (Bezos Type 1/Type 2 决策 + premortem Klein + Kahneman 论 executive bias + RAPID/DACI + base rates + 情景规划 + 反思 premortem); (g) 对内对外沟通 (创始人 / CEO 致股东信作为机构构建工具: Bezos / Buffett / Reed Hastings 备忘录 / 任正非 内部讲话 + Town Hall + 危机沟通 Tylenol/Boeing/Wells Fargo + 应对维权投资人); (h) 跨阶段规模化 (创业 0→1 / 扩张 1→100 / 成熟期掌舵 / 危机转型): Steve Blank + Brian Chesky founder-mode + Marc Andreessen + Pierre Lassonde + Lou Gerstner 转型; (i) 创始人 CEO vs 职业经理人 CEO 的差异 (Paul Graham 「Founder Mode」2024 + Brian Chesky + Reed Hastings (创始人→职业过渡) + BCG/McKinsey 的「manager mode」+ 两者各自代价 + 什么时候切换)。诚实处理: 幸存者偏差 (单个 CEO 的打法 ≠ 普适规律, 学术研究显示 CEO 个人方差约 ⅓, 约 70% 来自行业 / 时代 / 运气 / 资本结构), 创始人崇拜, 区分可复制工艺与不可移植个人魅力, 治理与利益相关者责任 (员工 / 股东 / 社会), 反对 CEO 神话与 Welch 退休金扭曲 / Theranos / WeWork / FTX / Uber-Kalanick 时代 / Musk 80h+ / 996 工时崇拜。覆盖 founder-CEO (Bezos/Hastings/黄仁勋/Chesky/Brin/任正非) 与职业 CEO (Nadella/Gerstner/Pichai/Mulally/Lafley)。不含: 泛泛管理理论 (商业书已覆盖, 本 skill 专谈 C-suite 这个具体岗位), 不含: 「领导力」作为个人成长自助 (Robbins/Tracy/Maxwell 励志体裁), 不含: 「如何创业」早期战术 — 一旦有员工 / 董事会 / 资本, CEO 的工艺才开始。 (CEO (Chief Executive Officer) — the craft of being the top operator of a company, distinct from generic 'leadership/management': (a) setting the strategy and crystallizing the vision (Drucker's 'theory of the business' + Porter on positioning + Hamilton Helmer's 7 powers + Andy Grove's strategic inflection points + Roger Martin Playing to Win); (b) capital allocation as the CEO's single highest-leverage job (Buffett/Munger Berkshire letters + William Thorndike 'The Outsiders' on 8 unconventional CEOs + Michael Mauboussin on hurdle rates + Bezos's annual shareholder letter approach to invariants vs bets + Mark Leonard Constellation Software); (c) building and leading the executive team (Patrick Lencioni 'Five Dysfunctions' + Ben Horowitz 'Hard Things About Hard Things' + Reed Hastings 'No Rules Rules' on talent density + Andy Grove HOP + Lou Gerstner IBM turnaround + Marshall Goldsmith); (d) board and investor relations (Larcker & Tayan 'Corporate Governance Matters' + Bill George 'True North' + the formal CEO-Chair / CEO-Lead Director protocols + activist investor playbooks + capital markets communication); (e) culture and values as operating system (Netflix culture deck + Amazon Leadership Principles + Bridgewater 'Principles' Ray Dalio + Schein on org culture + 任正非 Ren Zhengfei 灰度 + 张瑞敏 Zhang Ruimin 人单合一 + 稻盛和夫 阿米巴); (f) high-stakes decision-making under uncertainty (Bezos Type 1 vs Type 2 + premortem Klein + Kahneman 'Thinking Fast and Slow' for executive bias + RAPID/DACI + base rates + scenario planning + premortem); (g) internal and external communication (founder/CEO letters as institution-builders: Bezos shareholder letters / Buffett letters / Reed Hastings memos / 任正非 内部讲话 + Town Halls + crisis communication Tylenol/Boeing/Wells Fargo + activist responses); (h) cross-stage scaling (founder-CEO of 0→1 / scaling 1→100 / mature-stage stewardship / turnaround): Steve Blank + Brian Chesky founder-mode + Marc Andreessen + Pierre Lassonde mining vs steady-state + Lou Gerstner turnaround; (i) the founder-CEO vs professional-manager-CEO distinction (Paul Graham 'Founder Mod

日本語の概要は準備中です。原文の説明を表示しています。

Check an educational practice or task for cultural bias, WEIRD assumptions, and cross-cultural validity. Use when adapting resources for diverse contexts or questioning universal claims.

日本語の概要は準備中です。原文の説明を表示しています。

Physical principles underlying diodes, bipolar junction transistors, and MOSFETs — carrier concentrations, drift and diffusion, the pn junction, forward and reverse bias, the Shockley diode equation, BJT operating regions, MOSFET threshold and saturation, small-signal models, and temperature effects. Use when reasoning about device-level behavior, designing bias networks, selecting transistor operating points, interpreting datasheets, or diagnosing nonlinear device failures.

日本語の概要は準備中です。原文の説明を表示しています。

Use before any research activity or significant decision. Reviews cognitive biases relevant to the current stage.

日本語の概要は準備中です。原文の説明を表示しています。

Conduct a peer review of research methodology, experimental design, and manuscript quality. Covers methodology evaluation, statistical appropriateness, reproducibility assessment, bias identification, and constructive feedback. Use when reviewing a manuscript, preprint, or internal research report, evaluating a research proposal or study protocol, assessing evidence quality behind a claim, or reviewing a thesis chapter or dissertation section.

日本語の概要は準備中です。原文の説明を表示しています。

Run the memex session-init ritual at the start of a fresh session: load the bias-log, read the persistent-self trail in order, run the verification scoreboard, and surface the next slice from the docs. Use at the very start of any session that operates on the memex repo, when reconstructing context before doing work, or when an interrupted session needs to re-orient. NOT the `memex init` CLI command (which initializes the store/db) — this is the reconstruct-context ritual.

日本語の概要は準備中です。原文の説明を表示しています。

Comprehensive primary skill for agents working with Weights & Biases. Covers both the W&B SDK (training runs, metrics, artifacts, sweeps) and the Weave SDK (GenAI traces, evaluations, scorers). Includes helper libraries, gotcha tables, and data analysis patterns. Use this skill whenever the user asks about W&B runs, Weave traces, evaluations, training metrics, loss curves, model comparisons, or any Weights & Biases data — even if they don't say "W&B" explicitly.

日本語の概要は準備中です。原文の説明を表示しています。

APSA style checker for Quarto (.qmd) files. Checks numbers, capitalization, abbreviations, italics, in-text citations, titles of works, neutral and unbiased language, and APSA-specific terminology against the APSA Style Manual for Political Science (2018, updated 2023). Produces a structured markdown report organized by document section — never modifies the source file. Use when asked to check APSA style, fix citations, review capitalization, check number formatting, or flag biased language in a .qmd document. For grammar, spelling, and punctuation, use the proofread skill instead. Supports an optional output-file argument and an optional @sec-label argument to restrict checking to one section.

日本語の概要は準備中です。原文の説明を表示しています。

"Apply meta-analysis to synthesize effect sizes across multiple studies, assess heterogeneity, and evaluate publication bias. Use this skill when the user needs to combine findings from prior research, compare fixed-effect vs random-effects models, compute pooled effect sizes, or when they ask 'what does the overall evidence say', 'how do I combine results across studies', or 'is there publication bias'.".

日本語の概要は準備中です。原文の説明を表示しています。

"Apply AI ethics frameworks (fairness, accountability, transparency, privacy) to evaluate AI systems for algorithmic bias, explainability gaps, and value alignment failures. Use this skill when the user needs to audit an AI system for ethical risks, design fairness constraints, assess explainability requirements, or when they ask 'is this AI system fair', 'how do we detect algorithmic bias', 'what are the ethical implications of this AI deployment', or 'how do we make this model explainable to stakeholders'.".

日本語の概要は準備中です。原文の説明を表示しています。

Audit legal case prediction systems for bias, fairness, accuracy, and ethical guardrails. Use when: 'check my prediction model for bias', 'audit case outcome fairness', 'evaluate legal ML model', 'review sentencing prediction ethics', 'analyze bail risk algorithm', 'fairness metrics for justice system AI'.

日本語の概要は準備中です。原文の説明を表示しています。

Normalizes and transforms RNA-seq count matrices for DE, visualization, clustering, and ML. Covers between-sample (TMM, TMMwsp, RLE/median-of-ratios, upper quartile), within-sample (TPM, FPKM/RPKM), variance-stabilizing (VST, rlog, log-CPM), GC-content correction (cqn, EDASeq), and single-cell (scran deconvolution, scanpy normalize_total). Encodes the composition-bias rationale, the "most genes not DE" assumption and its catastrophic failure modes (MYC amplification, apoptosis, viral host shutoff, prokaryotic stress), the "lengthScaledTPM is not TPM" naming trap, the "TPM is not for DE" rule, the blind=TRUE vs FALSE decision, ERCC spike-in normalization (SBN), and the single-cell zero-inflation breakdown of TMM/RLE. Use when choosing or applying normalization, debugging shifted-MA-plot diagnostics, handling zero-heavy single-cell data, or correcting GC bias.

日本語の概要は準備中です。原文の説明を表示しています。

Normalize read-depth copy-ratio profiles and segment them into copy-number regions using circular binary segmentation (CBS, DNAcopy), hidden Markov models, HaarSeg, and fused-lasso methods. Covers GC-content, mappability, and replication-timing (wave-artifact) bias correction, panel-of-normals/PCA denoising, diploid-baseline centering, and algorithm selection by sequencing depth and event size. Use when choosing a segmentation algorithm, correcting depth bias, diagnosing oversegmentation or a mis-centered baseline, tuning CBS or HMM parameters, or understanding why a downstream CNV caller produced fragmented or shifted segments.

日本語の概要は準備中です。原文の説明を表示しています。

Detects allele-specific transcription factor or histone modification binding from heterozygous-variant ChIP-seq using WASP (reference-bias filter; mandatory upstream), RASQUAL (joint QTL + bias-corrected testing), BaalChIP (Bayesian beta-binomial with copy-number-aware overdispersion), and AlleleSeq (personalized diploid genome). Handles imprinted-locus awareness, X-inactivation artifacts, cancer copy-number imbalance, and integration with downstream caQTL / bQTL mapping. Use when identifying variants with allelic effects on TF binding, fine-mapping causal regulatory variants, validating deep-learning variant predictions, or characterizing cis-acting regulatory effects.

日本語の概要は準備中です。原文の説明を表示しています。

Identifies differentially bound ChIP-seq regions between conditions using DiffBind, csaw (sliding windows), DESeq2/edgeR/PyDESeq2 on count matrices, NormR (control-aware), or MAnorm2. Distinguishes three distinct normalization problems (composition bias, trended bias, global shifts) and matches each to its appropriate fix including spike-in scaling. Use when comparing ChIP-seq binding between experimental conditions, choosing normalization for global vs local changes, integrating spike-in data, or reconciling DiffBind/DESeq2 disagreement.

日本語の概要は準備中です。原文の説明を表示しています。

Estimates bivariate genetic correlation (rg) between traits from GWAS summary statistics or individual-level genotypes using cross-trait LDSC, HDL, LAVA, rho-HESS, GREML-bivariate, Popcorn, and HDL-L. Use when quantifying shared genetic architecture between two traits, screening MR validity before causal inference, distinguishing global from locus-level rg, estimating trans-ancestry rg, separating partial from full causation via LCV gcp, or producing a STROBE-MR-compliant cross-trait sensitivity battery. Cross-trait LDSC intercept absorbs sample overlap and is NOT a bias; HDL is biased under sample overlap above ~5%. High rg between exposure and outcome motivates CHP-aware MR sensitivity (CAUSE, LHC-MR).

日本語の概要は準備中です。原文の説明を表示しています。

Detect transcription factor binding footprints in ATAC-seq using TOBIAS, HINT-ATAC, Wellington, or scprinter. Use when identifying bound TF sites within accessible regions, correcting Tn5 insertion bias before footprinting, choosing between cleavage-based and aggregate-based footprinters, or comparing differential TF activity between conditions.

日本語の概要は準備中です。原文の説明を表示しています。

This skill should be used when the user asks to "implement LLM-as-judge", "compare model outputs", "create evaluation rubrics", "mitigate evaluation bias", or mentions direct scoring, pairwise comparison, position bias, evaluation pipelines, or automated quality assessment.

日本語の概要は準備中です。原文の説明を表示しています。

Rastreie experimentos de ML com logging automático, visualize treinamento em tempo real, otimize hiperparâmetros com sweeps e gerencie registro de modelos com W&B - plataforma colaborativa de MLOps

日本語の概要は準備中です。原文の説明を表示しています。

This skill should be used when the user asks to "implement LLM-as-judge", "compare model outputs", "create evaluation rubrics", "mitigate evaluation bias", or mentions direct scoring, pairwise comparison, position bias, evaluation pipelines, or automated quality assessment.

日本語の概要は準備中です。原文の説明を表示しています。