Search papers, labs, and topics across Lattice.
The paper introduces Pangu-ACE, a cascaded expert system for educational response generation on the EduBench benchmark, utilizing a 1B parameter tutor-router to generate initial responses and routing signals, escalating to a 7B specialist model when necessary. This approach improves deterministic quality from 0.457 to 0.538 and format validity from 0.707 to 0.866 compared to a legacy system, while handling nearly 20% of requests with the smaller model. Task-dependent routing reveals that simpler tasks are often handled by the 1B model, while more complex tasks require the 7B specialist.
A cascaded expert system can significantly improve the quality and format validity of educational response generation, but realizing latency gains remains a challenge.
Educational assistants should spend more computation only when the task needs it. This paper rewrites our earlier draft around the system that was actually implemented and archived in the repository: a sample-level 1B to 7B cascade for the shared-8 EduBench benchmark. The final system, Pangu-ACE, uses a 1B tutor-router to produce a draft answer plus routing signals, then either accepts the draft or escalates the sample to a 7B specialist prompt. We also correct a major offline evaluation bug: earlier summaries over-credited some open-form outputs that only satisfied superficial format checks. After CPU-side rescoring from saved prediction JSONL, the full Chinese test archive (7013 samples) shows that cascade_final improves deterministic quality from 0.457 to 0.538 and format validity from 0.707 to 0.866 over the legacy rule_v2 system while accepting 19.7% of requests directly at 1B. Routing is strongly task dependent: IP is accepted by 1B 78.0% of the time, while QG and EC still escalate almost always. The current archived deployment does not yet show latency gains, so the defensible efficiency story is routing selectivity rather than wall-clock speedup. We also package a reproducible artifact-first paper workflow and clarify the remaining external-baseline gap: GPT-5.4 re-judging is implemented locally, but the configured provider endpoint and key are invalid, so final sampled-baseline alignment with GPT-5.4 remains pending infrastructure repair.
