一种可自我改进的智能体,能生成自身的领域特定语言(DSL):用于精确问答的无监督算子归纳
A Self-Improving Agent that Grows its Own DSL: Oracle-Free Operator Induction for Exact Question Answering
AI总结:
该研究提出一种无监督的自改进智能体,通过嵌套循环在DSL中归纳算子,无需奖励、评判模型或真实标签,可扩展精确问答能力,在多领域及跨模型家族均表现出良好效果。
摘要:
可自我改进的智能体必须决定保留哪些新技能,在实践中,它会保留任何能获得高奖励、说服评判模型或压缩其库的技能,即便该技能是错误的。这在答案必须完全正确时就会出问题:在交易表中,总和要么正确要么错误。我们提出一个更具挑战性的问题:智能体能否在无奖励、无评判模型、无答案密钥的情况下,扩展自身的计算能力,且仍能保证答案完全正确?一个单一的冻结模型在小型、精确的领域特定语言(DSL)上运行两个嵌套循环。内层循环仅当两个独立编写的表达式在每条记录上都一致时,才回答新问题。外层循环会发现自身缺失的能力并为其编写算子,该算子仅在独立的第二个实现达成一致且结果符合守恒定律时才被接纳,全程无需真实标签。内层循环使可回答问题集大致增加两倍,外层循环在不相交的起始集上自我扩展,发明超出起始闭包范围的算子;无归纳的对照组则无法回答任何新问题。接纳规则经实验验证是可靠的:在独立参考下,它拒绝了138个有缺陷的算子变体(包括由不同模型家族编写的变体),而该领域依赖的“只要能运行就接纳”、自一致性和压缩信号则接纳了其中大部分。该机制无需修改即可在另外两个具有不相关模式的领域(商业账本和临床就诊问答)上自我扩展,使它们的可回答集分别增长十倍以上和近二十倍。该循环也可跨模型家族迁移:与其他提议者一起运行时,门控机制保持可靠,不同模型家族的自我扩展效果相当,仅归纳产量随每个提议者的DSL流畅度变化。
英文摘要:
A self-improving agent has to decide which new skills to keep, and in practice it keeps whatever earns a high reward, convinces a judge model, or compresses its library, even when the skill is wrong. That is fine until the answers have to be exactly right: over a table of transactions, a total is either right or wrong. We ask a harder question: can an agent grow what it can compute with no reward, no judge, and no answer key, and still be exactly right? A single frozen model runs two nested loops over a small, exact domain-specific language (DSL). The inner loop answers new questions only when two independently written expressions agree on every record. The outer loop discovers its own missing capability and authors an operator for it, admitted only when an independent second implementation agrees and the result obeys a conservation law, all without ground truth. The inner loop roughly triples the answerable-question set, and the outer loop self-extends on disjoint starting sets, inventing operators outside the starting closure; a no-induction control answers none of the new questions. The admission rule is empirically sound: it rejects 138 of 138 buggy operator variants under an independent reference, even one written by a different model family, where the admit-if-it-runs, self-consistency, and compression signals the field relies on admit most of them. Unmodified, the same mechanism self-extends on two further domains with unrelated schemas, business-ledger and clinical-encounter question answering, growing their answerable sets more than tenfold and nearly twentyfold. The loop transfers across model families too: run with further proposers, the gate stays sound and independent families self-extend comparably, with only the induction yield tracking each proposer's DSL fluency.
机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。