Cite
Notes
Only stored in your browser.
Attribution
RL environment for the mathematics of deep learning — information theory, backpropagation, optimisation, ML linear algebra and numerical stability ...
RL environment for Python packaging semantics — PEP 440 versions and specifiers, PEP 508 markers, extras, and dependency resolution with backtracki...
RL environment for exact linear algebra — RREF, the four subspaces, determinants, LU and Gram-Schmidt, eigenvalues, projection and least squares — ...
RL environment for discrete mathematics — recurrences, sums, integer functions, binomial coefficients, Stirling/Bell/Eulerian/Catalan numbers and e...
RL environment for machine-level computer systems — two's complement, integer overflow, IEEE-754, bit manipulation, struct layout and caches — with...
RL environment for FastAPI routing, request validation, response models and dependencies, graded by exact comparison of real HTTP responses.
RL environment for BeautifulSoup HTML extraction (selection, attributes, navigation, tables, malformed markup), graded by exact row comparison.
RL environment for JAX array, autodiff, linalg, nn and PRNG tasks, graded by deterministic allclose comparison.
RL environment for numpy/scipy numerical tasks, graded by deterministic assert_allclose comparison.
RL environment for pandas DataFrame tasks, graded by deterministic assert_frame_equal comparison.
RL environment that PLAYS single-player TextArena games to completion and takes the game's own verdict as the reward — seeded for determinism, and ...
RL environment for multi-hop composition over a synthetic fact universe — two- and three-hop chains, distractors, broken chains and aggregates — wh...
RL environment for automata and formal languages — DFA/NFA simulation, subset construction, minimisation, regular expressions, CYK parsing and Turi...
RL environment for classical AI — uninformed and informed search, alpha-beta pruning, constraint satisfaction, exact probability and MDPs — with st...
RL environment for the quantitative side of scaling ML models — rooflines, transformer FLOPs, KV-cache and optimizer memory, collective communicati...
RL environment for Django ORM tasks (filtering, relations, aggregation, expressions), graded by exact row comparison.
RL environment for the verifiers library itself — parsers, reward functions and Rubric composition, graded with no model in the loop.
RL environment for Matplotlib/Seaborn plotting, graded by reading values back off the axes rather than comparing images.
RL environment for scikit-learn preprocessing, modelling, metrics and validation tasks, graded by deterministic array comparison.
RL environment for PyTorch tensor, autograd, nn, loss and optimiser tasks, graded by deterministic torch.allclose comparison.
RL environment for exact pattern recognition — Bayes and conjugate priors, entropy and KL, least squares and ridge, naive Bayes, Bernoulli mixtures...