THE SOTA LAB · COMPUTATIONAL INSTRUMENTS & BENCHMARKS

Interactive Systems

We build to understand. Our research systems are explorable instruments that expose transformations from input to representation to inference.

Instrument 01

Bayesian Capability Inference Kernel

Conjugate Beta-Binomial online state filter with non-stationary evidence decay. Observe how the posterior density sharpens its bounds as observations accumulate.

Bayesian Belief State P(θ | E)Target Competency: Python / Data Engineering
Latent Expectation E[θ]
68%90% Credible Interval: [45%, 91%]
α = 6.8 (positive evidence)
β = 3.2 (counter evidence)
0.0 (Uncertain)0.5 (Prior)1.0 (Mastery)
Inject Observable Evidence:

Research insight: Every additional observation updates the posterior belief without asserting unjustified certainty. As α and β grow, the variance shrinks, narrowing the [45%, 91%] interval.

Complexity
O(1) Online Update

Constant-time algebraic conjugate parameter updates with zero iterative sampling overhead.

Drift Attenuation
γ ∈ [0.92, 0.99]

Mitigates score lock-in by decaying historical pseudo-counts over idle intervals.

Interface
JSON-RPC / REST

Standardized schema returning mean θ, variance σ², and 90% credible bounds.

Open Science

Reproducible Datasets & Benchmarks

Synthetic Drift Benchmark (N=10,000)CC-BY-4.0

Synthetic practitioner trajectories featuring career transitions, domain shocks, and burst observation periods.

OFO & ESCO Alignment GraphOpen Data

Cross-mapped competency vectors aligning the South African Department of Higher Education OFO gazette with European ESCO occupational frameworks.