In simulation the router matches the empirically optimal subject strategy in 8 of 13 subjects, but retreats to a safe strategy where gains are marginal¶

Claim · All claims
Evidence · 1 study (2 entries) · causal r2 · q2
Currency · newest study 2026 · not yet checked against its sources · not reviewed by a person · recheck by 2029: tool and model accuracy changes quickly

Subclaims¶

q2 i? In 8 of 13 evaluated subjects the router's predicted prompt exactly matched the empirically optimal prompt (Gap = 0.000), with an overall alignment rate (Gap≤0.01) of 69.2%. → Po-Chin Chang 2026 q2 i? In History the router converged to Analytical_opt instead of the empirically superior Best strategy, reflecting risk-avoiding behavior when potential gains are marginal (<0.013). → Po-Chin Chang 2026 (2)

Evidence¶

Po-Chin Chang 2026¶

Po-Chin Chang, Nicholas Hogan, Aske Plaat, Michiel T. van der Meer. (2026). Learning to Prompt: Improving Student Engagement with Adaptive LLM-based High-School Tutoring. https://arxiv.org/abs/2606.20138 doi:10.48550/arxiv.2606.20138

q2 · i? · causal · r2

Empirical alignment analysis in simulation (Table 1) comparing the model's final predicted prompt against the maximum-average-score prompt per subject; non-matching subjects show regret margins averaging under 0.02.

"In 8 out of the 13 evaluated subjects, the routing model achieved a perfect match (Gap = 0.000), successfully adopting the statistically optimal pedagogical strategy."

Po-Chin Chang 2026 (2)¶

Po-Chin Chang, Nicholas Hogan, Aske Plaat, Michiel T. van der Meer. (2026). Learning to Prompt: Improving Student Engagement with Adaptive LLM-based High-School Tutoring. https://arxiv.org/abs/2606.20138

q2 · i? · causal · r2

Simulation routing-dynamics analysis (Figure 4b, Table 1): in History the model settles on Analytical_opt rather than Best, which the authors attribute to prioritizing "score stability over exploration" when gains are marginal (<0.013).

"However, in subjects like History (see Table 1 and Figure 4b), the routing model converges to Analytical_opt, instead of the empirically superior Beststrategy."

Discussion¶

Related Claims¶

• In simulation, an adaptive prompt router outperforms two static prompt baselines on pedagogical score across 390 scenarios — related
• The adaptive router reduces turns in long conversations (turns≥5) but not across the unfiltered dataset — related
• Live feedback shifts the router's strategy from the Feynman prompt favored in simulation to the scaffolding Coach prompt, with superior live performance — related