「bias」の検索結果

A short self-bias checklist to run at the START of any investment research task (stock screen / sector study / company deep-dive). Four biases that systematically warp AI research — leader-bias (only big caps), English-bias (miss JP/KR/TW players), narrative-bias (chase concept labels), confirmation-bias (only bullish evidence) — plus recency-bias. Load this first, then research with the corrections in mind. Not a workflow, just a 60-second attitude reset that materially improves coverage and honesty.

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。

Three-layer bias analysis on any URL, file, or text — auto-fetches the content and any cited study, then audits data-level biases, source conflicts of interest, and journalism-added distortions, separating what the data supports from what's editorialized. USE WHEN bias analysis, analyze bias, bias check, check this study, who funded this, is this source biased, fact-check article, methodological flaws, source credibility, what's wrong with this claim. NOT FOR psychological author analysis, research synthesis (use Research), entity due diligence.

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。

Assesses AI bias risks for GDPR Art. 9 special category data and AI Act Art. 10 data governance. Covers fairness metrics, bias detection methods, mitigation strategies, and documentation requirements for protected characteristics. Keywords: AI bias, special category, fairness metrics, discrimination, Art. 9, Art. 10.

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。

临床诊断思维 (临床诊断思维 / 临床推理 (Clinical Diagnostic Reasoning) — 医生『怎么想病』的元学科:从症状/体征/检查到诊断结论的认知操作系统,从业者(临床医生/住院医/规培生,尤其全科/急诊/大内科等以未分化主诉为生的科室)、医学生与医学教育者、诊断安全与质量改进研究者、以及做医疗 AI 辅助诊断产品的人的视角。覆盖: (a) 第一性张力 — **直觉模式识别 (System 1: illness scripts 疾病脚本 / pattern recognition / gestalt, 『资深人一眼认出 aunt Minnie』) ⇄ 分析性推理 (System 2: hypothetico-deductive 假设演绎 / Bayesian 概率更新)** 的 dual-process 双过程理论 (Kahneman→Croskerry/Norman 谱系), 资深≠更会分析而是『脚本库更大+校准更好』; 更深层论战 — **『认知去偏可教 (Croskerry: bias awareness / cognitive forcing strategies / diagnostic timeout) ⇄ 偏倚标签是马后炮、知识结构才是主因 (Norman/Sherbino/Monteiro: debiasing 干预 transfer 证据弱, bias 是 hindsight 标签)』** — 本行最核心学术对垒; 『概率思维 (验前概率 × 似然比 → 验后概率, Pauker-Kassirer test/treatment threshold 阈值模型) ⇄ 穷尽式排查 (rule-out everything / 防御性医疗 / VOMIT)』; 『临床 gestalt ⇄ 结构化临床决策规则 (Wells/PERC/HEART)』; 『诊断简约 Occam's razor ⇄ Hickam's dictum (病人可以同时得 N 个病)』; 『床旁体格检查复兴 (Verghese Stanford 25 / McGee 循证体检) ⇄ 影像检验替代床旁』; (b) 方法论正典 — illness script theory (Schmidt/Boshuizen: enabling conditions/fault/consequences 三段结构), problem representation + semantic qualifiers 问题表征与语义限定词 (Bowen NEJM 2006, 把病人翻成 one-liner), hypothetico-deductive model (Elstein 1978《Medical Problem Solving》: 早期假设生成+定向检验), Bayesian 工具箱 (sensitivity/specificity/LR/Fagan nomogram/SnNout-SpPin), threshold model (Pauker-Kassirer NEJM 1975/1980), 认知偏倚分类学 (anchoring/premature closure/availability/confirmation/base-rate neglect/search satisficing/diagnostic momentum — Croskerry 偏倚清单), 去偏与元认知策略 (diagnostic timeout/cognitive forcing/Ely checklist/calibration), schema-based reasoning (Clinical Problem Solvers schemas: 按 pivot 症状走分支), reflective practice 结构化反思 (Mamede/Schmidt), 诊断错误科学 (NAM 2015 报告定义 / Newman-Toker Big Three / Hardeep Singh e-triggers / SAFER Dx 框架); (c) 行业结构与角色 — 医学生→实习/住院医 (晨会 morning report/查房被 pimping/汇报训练)→主治 attending→master clinician (NEJM CPC discussant/晨会大师如 Dhaliwal); 配套生态: 医学教育者 (clinical reasoning curriculum + 评估: script concordance test/key features exam/OSCE), 诊断安全研究者 (SIDM/AHRQ), 诊断辅助与 CDS 工具开发者 (DDx generator/AI); (d) 核心工作流 — 数据采集 (病史为王 + 循证体检 + 针对性检验影像) → 问题表征 (one-liner + semantic qualifiers) → 鉴别诊断生成 (schema / 解剖定位法 / VINDICATE-M 病因筛, 按『常见可能 × 致命不能漏 can't-miss』双轴排序) → 假设定向检验 (按 LR 选检查 / 阈值决策) → working diagnosis + 安全网 (red flags 交代 / test of time / test of treatment) → 反馈校准 (follow-up / M&M / diagnostic timeout); 教学工作流: 晨会渐进披露汇报 / CPC / SNAPPS / one-minute preceptor / virtual morning report; AI 增强工作流 (2023-2026: LLM 鉴别诊断头脑风暴 / OpenEvidence 检索 / ambient scribe 释放认知带宽 + 自动化偏倚 guardrails); (e) 产出物 — one-liner, problem list, prioritized DDx, assessment & plan (按问题分层), 晨会/CPC 汇报, M&M 复盘, 诊断不确定性沟通 (『最可能是 X, 但出现 Y 红旗立刻回来』); (f) 教育与评估 — script concordance test / key feature exam / EPA / 里程碑; 中国语境: 人卫《诊断学》教材 / 执业医师考试 / 规培结业临床思维考核; (g) 争议/批判 — debiasing 之争 (Croskerry vs Norman-Sherbino『knowledge is the cure』, 去偏 RCT 效果弱 / bias 标签不可证伪), 诊断错误率数字之争 (『10-15%』经典估计 / Newman-Toker 79.5 万美国年严重伤害外推方法被质疑 / 尸检符合率), dual-process 二分被批过度简化 (连续谱 / 难以实证分离), 决策规则 vs gestalt (资深 gestalt 常不输 Wells 类规则 / 算法厌恶), vignette 研究外推性 + context/case specificity (推理高度内容绑定不可通用迁移 [Norman/Eva] — 对『教推理通用课』产业的根本批判), AI 辅助诊断 2023-2026 (GPT-4 在 NEJM CPC/vignette 追平或超医生 [Kanjee 2023 JAMA / Goh 2024 JAMA Netw Open『AI alone > physician+AI』悖论] / Google AMIE / OpenEvidence 爆发 vs 自动化偏倚 / 去技能化 / 真实环境验证缺失), 防御性医疗与过度检查 (incidentaloma 瀑布 / VOMIT), 体检衰亡之争, pimping 教学法争议; (h) 流派/思想谱系 — 决策科学/Bayesian 派 (Ledley-Lusted 1959 → Elstein/Kassirer/Sox/Pauker/Eddy → Brush) vs 认知心理/双过程派 (Kahneman-Tversky → Croskerry/Graber) vs 教育认知派 (Schmidt/Boshuizen illness scripts → Norman/Eva/Mamede/Durning 情境认知) vs 诊断安全/系统派 (NAM 2015 / Newman-Toker / Hardeep Singh / Graber-SIDM: 错误=系统×认知共因) vs 床旁临床派 (Osler 传统 → Tierney aphorisms / Verghese / McGee / Dhaliwal / Saint: 病史体检为王 + 刻意练习) vs AI/计算派 (INTERNIST-1/DXplain 专家系统 → Isabel → LLM 世代 Rodman/Topol)。不含: 具体专科治疗方案与手术决策深度 (足踝外科/种植牙等另有 skill), 中医辨证 (另有 skill), EBM 文献批判性评价方法学 (相关但本 skill 聚焦诊断推理本身), 护理诊断体系 (NANDA), 影像/病理知觉型读片训练 (相邻但独立), 精神科 DSM 结构化访谈细节, 医患沟通技巧全集 (只覆盖诊断不确定性沟通), 患者自查/自我诊断指

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。

Recognition and mitigation of systematic reasoning errors documented in the heuristics and biases tradition. Covers confirmation bias, availability, anchoring, representativeness, framing effects, hindsight bias, overconfidence, and motivated reasoning. Each bias is presented with its mechanism, diagnostic signal, and structured mitigation. Use when you suspect your own or another reasoner's conclusions may be shaped by systematic cognitive distortion rather than evidence.

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。

Azure Weights & Biases SDK for .NET. ML experiment tracking and model management via Azure Marketplace. Use for creating W&B instances, managing SSO, marketplace integration, and ML observability. Triggers: "Weights and Biases", "W&B", "WeightsAndBiases", "ML experiment tracking", "model registry", "experiment management", "wandb".

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。

Plans confounder control, variable adjustment logic, and bias mitigation strategies at the protocol stage for clinical, epidemiologic, translational, observational, and biomarker studies. Always use this skill when a user needs to identify major confounders, decide which variables should or should not be adjusted for, compare matching/stratification/weighting approaches, anticipate selection or measurement bias, or pressure-test a study design before execution. Focus on bias sensing, causal structure awareness, variable-role classification, and critical design review rather than generic statistical advice.

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。

Calls significant loops from protein-directed and targeted 3C assays (HiChIP, PLAC-seq, Capture Hi-C/PCHi-C, ChIA-PET) where the contact background is peak-anchored and coverage-biased, so generic Hi-C loop callers (cooltools dots, Juicer HiCCUPS) use the wrong null. Covers FitHiChIP (config-driven coverage+distance-decay spline regression, peak-to-peak vs peak-to-all foreground, loose vs stringent background, coverage vs ICE bias), MAPS (positive Poisson regression on bias factors for PLAC-seq/HiChIP), hichipper (restriction-site-distance bias model + library QC), CHiCAGO (Delaporte two-component Brownian+technical background for asymmetric bait x other-end Capture Hi-C), the with/without separate-ChIP anchor decision, and differential loops via diffloop. Use when calling loops from HiChIP/PLAC-seq/Capture Hi-C, choosing FitHiChIP/MAPS/CHiCAGO, picking peak-to-all vs peak-to-peak, setting the loop FDR, supplying ChIP peaks as anchors, QCing a HiChIP library, or comparing loops 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。

Diagnoses bias, anomalies, and strange-looking results on a specific PostHog experiment. Covers empty / 0-exposure experiments, sample ratio mismatch, identity fragmentation, multi-variant exposure, uneven-split exclusion bias, significance traps (peeking, A/A, Bayesian vs Frequentist), PostHog-vs-SQL discrepancies, and surprises after mid-run edits. Symptom-driven dispatch to the right diagnostic. TRIGGER when: user asks 'is my experiment biased?' or 'why 0 exposures?', references the bias banner, says a variant looks strange / wrong / off, sees significance flipping, notices PostHog numbers disagreeing with their SQL, sees an A/A test showing significance, or reports surprises after mid-run edits. DO NOT TRIGGER when: creating a new experiment (use creating-experiments), only configuring rollout (use configuring-experiment-rollout) or metrics (use configuring-experiment-analytics), or only asking lifecycle questions (use managing-experiment-lifecycle).

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。

Calls significant loops from protein-directed and targeted 3C assays (HiChIP, PLAC-seq, Capture Hi-C/PCHi-C, ChIA-PET) where the contact background is peak-anchored and coverage-biased, so generic Hi-C loop callers (cooltools dots, Juicer HiCCUPS) use the wrong null. Covers FitHiChIP (config-driven coverage+distance-decay spline regression, peak-to-peak vs peak-to-all foreground, loose vs stringent background, coverage vs ICE bias), MAPS (positive Poisson regression on bias factors for PLAC-seq/HiChIP), hichipper (restriction-site-distance bias model + library QC), CHiCAGO (Delaporte two-component Brownian+technical background for asymmetric bait x other-end Capture Hi-C), the with/without separate-ChIP anchor decision, and differential loops via diffloop. Use when calling loops from HiChIP/PLAC-seq/Capture Hi-C, choosing FitHiChIP/MAPS/CHiCAGO, picking peak-to-all vs peak-to-peak, setting the loop FDR, supplying ChIP peaks as anchors, QCing a HiChIP library, or comparing loops 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。

Score LLM and agent outputs using LLM-as-judge techniques — direct scoring against rubrics or pairwise comparison between two outputs. Includes built-in bias mitigation for position bias, length bias, and self-enhancement bias. Load when the user asks to score an output, judge a response, evaluate against a rubric, compare two outputs, do direct scoring, run pairwise comparison, or says "rate this", "which response is better", "score this against the rubric", "judge this output", "LLM as judge this". Sub-skill of eval-output orchestrator.

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。

Calls significant loops from protein-directed and targeted 3C assays (HiChIP, PLAC-seq, Capture Hi-C/PCHi-C, ChIA-PET) where the contact background is peak-anchored and coverage-biased, so generic Hi-C loop callers (cooltools dots, Juicer HiCCUPS) use the wrong null. Covers FitHiChIP (config-driven coverage+distance-decay spline regression, peak-to-peak vs peak-to-all foreground, loose vs stringent background, coverage vs ICE bias), MAPS (positive Poisson regression on bias factors for PLAC-seq/HiChIP), hichipper (restriction-site-distance bias model + library QC), CHiCAGO (Delaporte two-component Brownian+technical background for asymmetric bait x other-end Capture Hi-C), the with/without separate-ChIP anchor decision, and differential loops via diffloop. Use when calling loops from HiChIP/PLAC-seq/Capture Hi-C, choosing FitHiChIP/MAPS/CHiCAGO, picking peak-to-all vs peak-to-peak, setting the loop FDR, supplying ChIP peaks as anchors, QCing a HiChIP library, or comparing loops 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。

Research Guardian - Ethics Advisory & Bias Detection across all research stages Enhanced VS 3-Phase process: Surface-level screening, deep contextual analysis, constructive recommendations Use when: reviewing research ethics, checking for bias, assessing trustworthiness, QRP screening Triggers: ethics review, IRB, bias detection, QRP, trustworthiness, research integrity, p-hacking, HARKing

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。

Generates the "Risk of Bias" results section for a meta-analysis based on assessment tables and statistics. Use when the user wants to draft the risk of bias analysis text from provided data tables.

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。

Decomposed, multi-criteria metric design for LLM pipelines. The metric IS the model — change the metric and the optimizer changes behavior. Decompose by default; bool during compile, float during eval; calibrate against human; mitigate judge bias. Search keywords: LLM-as-judge, llm as judge, eval metric, evaluation score, scoring function, rubric, RAGAS, G-Eval, judge bias, verbosity bias, how to evaluate LLM output.

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。

Classifies sensitive data in AI/ML training datasets including bias detection for Art. 9 categories, data card documentation, provenance tracking, and consent verification for model training. Keywords: AI training data, ML dataset, bias detection, data card, model training, Art 9, consent, GDPR 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。

Guides the combined DPIA and AI Act conformity assessment for AI systems processing personal data. Covers EDPB-EDPS Joint Opinion 5/2021, training data lawfulness under Art. 6 and Art. 9, Art. 22 automated decision-making, algorithmic bias detection, and NIST AI RMF MAP function. Keywords: AI privacy, DPIA, AI Act, algorithmic bias, automated decision-making, Art. 22, training data, NIST AI RMF.

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 whether a paper's EVALUATION DESIGN actually measures what it claims and whether its reporting is complete — the validity layer family D (experiment-forensics) cannot reach. Three patterns: train/test leakage means the reported score may not measure generalization (HP-EVAL-LEAKAGE — adopts the Kapoor & Narayanan 8-type / 3-category leakage taxonomy; the illegitimate-proxy / sampling-bias / pretraining-contamination subtypes hand off as needs_external_check, naming but NEVER running Oren-2023 exchangeability / Shi-2023 Min-K% / Golchin-2023 Time-Travel / BIG-bench canary); a load-bearing LLM judge is conflicted (same model/family as a compared system) or unvalidated (no human-agreement, no bias control) (HP-JUDGE-VALIDITY); a declared condition/metric is dropped or switched to favor the method, or 'best' is chosen with no held-out set (HP-SELECTIVE-REPORTING). Verdict-bearing at L0/L1 from the DESCRIBED protocol — NOT repo-gated like experiment-forensics; L2 only CONFIRMS against split/preprocessing/result files. A fresh cross-model reviewer (gpt-5.6-sol xhigh, read-only, fresh thread per pass) PROPOSES findings, each span-anchored to a ledger claim_id; tools/adjudicate_findings.py DECIDES the verdict. Leakage and under-reporting are usually HONEST methodological errors — every finding describes a discrepancy to CHECK, never an accusation. An LLM generating GROUND-TRUTH labels is HP-FAKE-GT (experiment-forensics) — routed there, not here. Emits eval-design-forensics.findings.json; computes NO verdict. Detect-only. Triggers: "eval design audit", "evaluation validity", "train/test leakage", "data leakage", "is the score measuring generalization", "LLM judge bias", "is the judge validated", "selective reporting", "cherry-picked results", "评估设计审计", "评测有效性", "数据泄漏", "训练测试集泄漏", "裁判模型有没有验证", "选择性报告".

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。

Bias detection and mitigation, fairness metrics, privacy frameworks, consent models, transparency requirements, and accountability structures for data science practice. Covers algorithmic bias sources, disparate impact testing, differential privacy, GDPR principles, model cards, datasheets for datasets, responsible AI frameworks, and the organizational governance needed to make ethics actionable. Use when auditing models for bias, designing privacy-preserving systems, establishing governance processes, or evaluating the social impact of data-driven decisions.

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。

Cross-session shared memory for agents. Use at session start to load the persistent bias-log via `mcp__memex__recent_observations`, mid-session to search prior context with `mcp__memex__search`, and immediately when a bias surfaces to log a new `observation` via `mcp__memex__add`. The six-layer reconstruction trail (raw → embedded → indexed → graphed → curated → reflexive) means memex stores facts AND the agent's own reasoning patterns across sessions. Prefer querying memex before re-deriving prior architectural decisions; re-derivation is itself evidence that the trail is incomplete.

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。

Log a bias-log / vipassana observation to the memex store the moment a reasoning pattern surfaces, mid-session or at close. Use when you catch a bias in your own reasoning (anchoring, confirmation, trust-the-summary), when a pace tell or verification gap appears, or when meditate/observe produces a reflection worth persisting across sessions. Primary path pipes the body to `memex add` over stdin (`--title` required) then runs the `meditate-vipassana` extractor to make it queryable; secondary path appends a numbered bullet to `docs/OBSERVATIONS.md`.

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 AI systems for bias with testing frameworks and mitigation strategies. TRIGGERS - Use when user needs help with ai-bias-auditor related tasks.

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 AI systems for bias and fairness. TRIGGERS - Use when user needs help with ai-bias-audit related tasks.

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 behavioral finance theory to identify systematic investor biases and their impact on asset prices. Use this skill when the user needs to analyze irrational market behavior, explain pricing anomalies through cognitive biases, diagnose investor decision errors, or when they ask 'why do investors hold losers too long', 'how does loss aversion affect pricing', or 'what biases drive this market pattern'.".

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。