EduClaw-Bench¶
Product or Programme · All products and programmes
Evidence · 2 claims (2 for) · 1 study (1 design),q2· 0 of 1 report an effect size · 2 claims rest on one study
Description¶
EduClaw-Bench is a virtual benchmark environment developed by Lee et al. for evaluating pedagogical LLM tutoring agents over simulated 30-day elementary-mathematics learner interactions.
Components¶
Claims¶
- Tutoring quality belongs to the base model and agent harness together rather than either alone, as adapter rankings reorder across tiers [+M]
- Almost no model-and-harness combination sustains tutoring over the full 30-day horizon, as learning plateaus within 5–10 days [+M]
Related Products and Programmes¶
Key Sources¶
- Lee, U.; Lee, S.; Jeong, Y.; Lee, E.; Shin, M.; Kwon, H. (2026). EduClaw-Bench: A Long-Horizon Benchmark for Pedagogical LLM Agents with Simulated Learners. https://arxiv.org/abs/2608.03206. doi:10.48550/arxiv.2608.03206
Page metadata
| Field | Value |
|---|---|
type |
product |
id |
educlaw-bench |
title |
EduClaw-Bench |
description |
EduClaw-Bench is a virtual benchmark environment developed by Lee et al. |
status |
review |
product_kind |
software |
generated |
by process:settle-candidates on 2026-10-10 |
trust |
unverified — no one has confirmed this page's content |
Frontmatter is the page's machine-readable record — see the Schema & Guide.