ALTK-Evolve 对比 ACE：按需检索降低推理成本并提升准确率
原标题：Thinking of ACE? We Can Do It with Fewer Tokens 17 August 11, 2026
AI 摘要
Hugging Face 博客介绍了 ALTK-Evolve 与 ACE 两种智能体记忆系统，均通过将智能体历史轨迹转化为可复用经验并在推理时注入，避免压缩经验。ALTK-Evolve 采用按需检索少量指南的方式，而 ACE 每步注入完整手册。在 AppWorld 基准上，ALTK-Evolve 在 DeepSeek-V3.2 上准确率更高且 token 消耗仅为 ACE 的约 40%，在 gpt-oss-120b 上准确率相近但成本约为七分之一。
正文节选
Give an LLM agent a realistic multi-step task — split a bill, find a song, reconcile an order across nine simulated apps — and when it fails, it usually isn't for lack of knowledge. It mis-paginates an API, resolves the wrong person, or returns a value when none was asked for. The model knows the APIs; what it hasn't internalized is how to use them reliably. That's learnable from the agent's own history. Two recent systems do exactly this, on the same kind of agent: ACE (Agentic Context Engineer