Ripple-65B Preview 发布：美国开源权重模型在智能体编码上达到 SOTA
Ripple-65B Preview: U.S. Open-Weight SOTA on Agentic Coding
ricursive 发布 Ripple-65B Preview，称其在美国开源权重模型中于智能体编码任务上达到 SOTA，TerminalBench 2.1 得分 80.5。团队未采购后训练数据，自建合成数据管线与质量过滤，混合 RL 配方使用截断重要性采样、分数中心化和线性长度惩罚，仅用 1.4M rollouts 完成训练，约为前一 SOTA Reflection Beam 的七十分之一。
Rather than investing in a brute-force RL climb, we carefully iterated on our SFT and RL recipes with a focus on compute efficiency. We opted not to buy any posttraining data, instead devising our own synthetic data pipelines, quality filtering, and coverage metrics.
Our mixed RL recipe uses
,
, and
with high-quality long-horizon synthetic environments. The RL tasks are tuned for our model capabilities, and we continuously added harder tasks as the model improved.
The end result was a steep RL climb from a strong initial SFT checkpoint. We reached a TerminalBench 2.1 score of 80.5 with only 1.4M rollouts, roughly 70x fewer rollouts than Reflection Beam, the prior SOTA.
来源：Hacker News · ricursive.com