Representative gridworld teaching tools lack live Bellman substitution, decision-provenance logging, and full trace export¶

Claim · All claims
Evidence · 1 study · design r2 · q1
Currency · newest study 2026 · not yet checked against its sources · not reviewed by a person

Subclaims¶

q1 i? Among the representative gridworld Q-learning teaching tools reviewed, none simultaneously renders the update equation with live numeric substitution, logs the full pre-decision Q-row with greedy/random provenance, and exports the complete decision trace for learner-side analysis. → Jueng 2026

Evidence¶

Jueng 2026¶

Jueng, Ekkachai. (2026). Q-Learning Lab: Teaching Reinforcement Learning Through Learner-Generated Trace Analysis. https://arxiv.org/abs/2607.10802 doi:10.48550/arxiv.2607.10802

q1 · i? · design · r2

Feature comparison against representative tools (CS188 Gridworld, RL Playground, Virtual Labs, tinyrl, gridworld-qlearning), summarized in Table 1 from publicly documented feature sets. The authors report they 'did not find a system that simultaneously' offers the three features.

"we did not find a system that simultaneously (i) renders the update equation itself with live numeric substitution at every step, (ii) logs the agent’s full pre-decision Q-row together with the greedy/random provenance of each action"

Discussion¶

Related Claims¶

• Of 100 human-audited claims, 39 are fully verifiable, 55 partially verifiable, and 6 not verifiable, with main responses showing stronger provenance than follow-ups — related