首页 > AI前沿 > Measure Learning at Steady State: A BIRD-SQL Formula 1 Case Study

Measure Learning at Steady State: A BIRD-SQL Formula 1 Case Study

arXiv机器学习 2026-09-11 20:01 6 阅读 查看原文

Continual Learning Bench scores learning as short-horizon gain versus a reset baseline and finds naive full-context ICL strongest among the memories it tested.

We treat ICL as one learning system and score it on a longer shared-world schedule.

Steady-state learning is the gap versus baseline on a pre-set late window (last 40 of 174 BIRD-SQL formula-1 questions).

We split the score into exploration efficiency (SQL probes), task reward (hits), and delivery cost (API dollars and context size).

On gpt-5.6-luna, late probes fall from 4.6-5.6 to 0.95 while hits rise only modestly and ICL context grows to about 95k tokens with cost roughly doubling.

Short-horizon gain understates the late probe saving and misses the cost inversion, so we find that unbounded ICL is a poor candidate for the learning mechanism.