Results for “depth-estimation”

2-stage pipeline: trace (causal investigation) -> deep-interview (requirements crystallization) with 3-point injection

Cross-runtime 2-stage pipeline for Claude Code, Codex/OMX, and Gemini/Antigravity/OMA: trace causal hypotheses, inject evidence into deep-interview style requirements crystallization, then hand off to the right runtime planner/executor.

Trains, evaluates, exports, and runs inference on FoundationStereo models for stereo depth estimation and 3D reconstruction from stereo image pairs.

Trains, evaluates, exports, and runs inference for FastFoundationStereo (FFS) stereo depth estimation models, a distilled variant of FoundationStereo with lower latency.

Identifies the weakest samples per ground-truth label in NVIDIA TAO VCN Classify experiments by running a Docker container that performs threshold sweep, weakness scoring, and per-lighting expansion, then surfaces top-K weak samples for downstream augmentation or relabeling.

Profile a DeepStream pipeline with Nsight Systems and derive its configs from the measurement.

選択肢(A/B/C)の深掘り比較→淘汰→推奨で判断負担を下げ判断の質を上げるスキル。5段階（L1固定3点/L1.5案拡張Diverge・自動/L2評価軸マトリクス/L3複数LLM弁証論/L4過去判断照合）。 「比較して」「深掘りして」「メリデメ教えて」「お勧めは？」「徹底的に」「過去の判断と照合」「前にどう決めたっけ」「/sentaku」等で発火。teian(浅)の深掘り要求を受け取り、brainstorming(深:設計全体)と棲み分け。

Detects DNS-based data exfiltration, tunneling, and DGA communication by analyzing query logs with entropy analysis, volume anomalies, and subdomain length detection in SIEM platforms.

Conducts digital forensics investigations on compromised endpoints, including memory acquisition, disk imaging, artifact analysis, and timeline reconstruction for incident response and evidence collection.

Expert in residential hollow space detection, hidden room discovery, and safe room planning. Helps map house dimensions, identify anomalies suggesting hidden spaces, and safely explore potential voids. Knowledge of architectural history, construction methods, and non-destructive investigation techniques. Activate on "panic room", "hidden room", "secret room", "hollow space", "house mapping", "find hidden space", "room dimensions", "hidden door", "false wall", "priest hole", "prohibition era", "safe room". NOT for illegal entry, structural modifications without permits, or bypassing security systems.

DeepEval — LLM evaluation framework, RAG metrics, hallucination detection, red-teaming, CI/CD integration

Performs privilege escalation assessments on compromised Linux and Windows systems to identify paths from low-privilege access to root or SYSTEM-level control.

Extracts GraphQL API schemas through introspection attacks, identifies sensitive fields and mutations, and tests for query depth and complexity vulnerabilities.

Provides a structured framework for analyzing cyber intrusions by examining four core features: Adversary, Capability, Infrastructure, and Victim. Covers implementing the Diamond Model programmatically to classify and correlate intrusion events, build activity threads, and generate pivot-ready intelligence.

Analyze user feedback data to identify segments with sentiment scores, JTBD, and product satisfaction insights.

Guides a structured interview and repository research to clarify feature requirements, classify scope, and produce context artifacts for approval before design.

Influence Functions in Deep Learning

Detect data exfiltration through DNS tunneling by analyzing query entropy, subdomain length, query volume, TXT record abuse, and response payload sizes using passive DNS monitoring.

Investigates insider threat incidents involving employees, contractors, or trusted partners who misuse authorized access to steal data, sabotage systems, or violate security policies. Combines digital forensics, user behavior analytics, and HR/legal coordination to build an evidence-based case.

Analyzing pre-change impact across vertical (dependency chains, files) and horizontal (pattern consistency, naming) dimensions. Use to estimate blast radius before a refactor. No code.

Quantifies audio source separation quality by computing the signal-to-distortion ratio (SDR) between ground-truth and estimated stems, with per-stem and record-level averaging.

Strategic patterns for solving intractable problems through cascading approximation, self-improvement, and heterogeneous evaluation from DeepMind's AlphaGo system

Scans a codebase for deepening opportunities, presents them as a visual HTML report, then grills through the chosen one. Use when the user wants to improve codebase architecture or find refactoring opportunities.

Comprehensive results analysis for empirical research: generate publication-quality descriptive statistics and balance tables, interpret regression coefficients with economic magnitude and effect sizes, assess identification assumption diagnostics, and produce structured results memos. Use when asked to create summary statistics, Table 1, balance tests, interpret results, assess economic significance, or write results narratives.

Detects insider data exfiltration by analyzing DLP policy violations, file access patterns, upload volume anomalies, and off-hours activity in endpoint and cloud logs using pandas for behavioral analytics and statistical baselines.

Predict molecular properties, train graph neural networks, and run drug discovery workflows using DeepChem's featurizers, models, and MoleculeNet benchmarks.

Analyse text complexity across quantitative, qualitative, and reader-task dimensions with scaffolding recommendations. Use when selecting texts, assessing readability, or planning reading support.

Exploratory data analysis (EDA) on a scientific data file — auto-detects the format, runs structure/quality/statistics checks, and writes a markdown EDA report with downstream recommendations. Use when asked to "explore", "analyze", "summarize", "profile", or "QC" a data file, or to understand its structure/content/quality before deciding what analysis to run. Covers tabular (.csv .tsv .xlsx .parquet), arrays (.npy .npz .hdf5 .h5 .mat .fits), sequence/genomics (.fasta .fastq .sam .bam .vcf .bed .gff .gtf .h5ad), microscopy (.tif .nd2 .czi .lif .ims .dcm .nii), spectroscopy/MS (.mzML .mzXML .mgf .fid .jdx), chemistry (.pdb .cif .mol .sdf .xyz .gro), and proteomics/metabolomics (.pepXML .mzid .mzTab). For zero-shot forecasting of a series use alterlab-timesfm; to create/configure a chunked cloud array store use alterlab-zarr. Part of the AlterLab Academic Skills suite.

Expert guidance for distributed training with DeepSpeed - ZeRO optimization stages, pipeline parallelism, FP16/BF16/FP8, 1-bit Adam, sparse attention

Run Mom Test–style customer-discovery interviews to validate or kill an unbuilt idea — generate a non-leading interview guide, conduct or coach the conversations, and synthesize signal vs compliments. Load when the user asks to do customer discovery, run problem interviews, validate an idea with users, run a Mom Test, design an interview guide, or says "talk to customers", "validate the problem", "interview users", "Mom Test this", "did real users want it", "synthesize my interviews", "I just talked to N people". Sub-skill of `venture-exploration`. Hard-bans "would you use this?", solution-pitching, friend/family-only ICP, and treating compliments as validation. Calls `secure-*` before synthesizing any pasted external transcripts.

Enumerating failure modes via pre-mortem analysis. Systematically identifies failure scenarios for plans, designs, and features, scoring them with RPN/AP. Does not write code.

Conduct in-depth, multi-step research on a given topic by decomposing queries, finding diverse sources, cross-referencing findings, and synthesizing a comprehensive report. Use when the user requests deep research or provides relevant inputs for this workflow.

Auto-mirror top Discovery traders on Hyperliquid. Scans top 50 traders, scores on PnL rank (35%), win rate (25%), consistency (20%), hold time (10%), drawdown (10%). Creates 2-5 mirror strategies with overlap checks. Daily rebalance with 2-day watch period before swaps. Use when setting up trader mirroring, copy trading, or portfolio auto-rebalancing based on Discovery leaderboard performance.

Map product assumptions by risk, confidence, evidence, and next validation step.