Labs
Every number in a lab is computed in your browser, precomputed by a script in this repository, or produced by a labelled deterministic simulation. Nothing is a canned model output.
- 5full labs
- 55interactive figures
- 60/60built
Labs5
- ML and metricsThreshold, imbalance and regularization against a confusion matrix, precision/recall and cost.Open lab
- ForecastingTrend, seasonality, noise, horizon and cutoff against baselines, backtest error and intervals.Open lab
- Neural networkShapes, forward pass, loss, a gradient step and training curves versus a logistic baseline.Open lab
- RetrievalChunking and lexical, dense (LSA) and hybrid retrieval on labelled queries. No generator is configured.Open lab
- Cost and qualityTransparent cost, latency and quality trade-offs from editable, clearly hypothetical inputs.Open lab
All interactive figures60 of 60 built
- Task lensday 1
- Availability timelineday 2day 24
- Natural-frequency Bayes gridday 3
- Sampling and significance labday 4
- Confounding and Simpson labday 5day 10
- EDA detectiveday 6
- Grain and join explorerday 7day 10
- Chart criticday 8
- Waterfall and driver treeday 8day 18
- Preprocessing leak labday 9
- OLS playgroundday 11
- Logistic regression labday 11day 13
- Overfitting sliderday 11day 15
- Tree labday 12
- Ensemble labday 12
- Imbalance and threshold labday 13
- Cluster labday 14
- Split and leakage labday 15
- Validation schemesday 15day 25
- Metric fightday 16
- Decision cost calculatorday 16day 26day 59
- Interpretability labday 17
- Error slicerday 18
- Drift monitorday 19
- Time series component builderday 21
- ACF and PACF explorerday 22
- Decomposition labday 22
- Baseline raceday 23
- ETS explorerday 23
- Lag feature labday 24
- Interval labday 26
- Reconciliation labday 26
- Forecast family chooserday 27
- Forecast vs causal scenario labday 28
- Tensor shape tracerday 31
- Activation and loss labday 32
- Backprop stepperday 33
- Training curves labday 34
- Convolution calculatorday 36
- RNN unrollday 37
- Embedding spaceday 38
- Attention labday 39
- Training memory calculatorday 40day 44
- TF-IDF vs dense retrievalday 41
- Next-token samplerday 42
- Adaptation chooserday 43
- LoRA calculatorday 44
- LLM evaluation labday 45
- Quality-diversity trade-offday 46
- Structured output validatorday 47
- Agent vs workflowday 49
- Retrieval evaluation labday 50day 51day 57
- LLMOps traceday 52
- Injection labday 53
- Model selection matrixday 60