Perplexity 用提示引导自蒸馏让模型从自身错误中学习
Perplexity 把自家模型改成会从错误里自我纠偏，A/B 测试里工具调用失败少了 21.2%，做法写得很细。
Perplexity 发布新研究，采用 hint-guided self-distillation（提示引导自蒸馏）方法对 Computer 模型做后训练，让模型从自身错误中学习。
在一次线上 A/B 测试中，较晚训练的 checkpoint 相较较早版本将工具调用失败率相对降低了 21.2%。
该方法面向减少智能体场景中的工具调用错误。
New research: We post-trained a Computer model to learn from its own errors using hint-guided self-distillation. In a live A/B test, a later trained checkpoint reduced tool-call failures by 21.2% relative to an earlier checkpoint.📊 18 ⚡
This report is categorized under AITOP's "Papers" desk, focusing on US frontier AI labs and Silicon Valley technology leaders. Within our cross-regional analytical framework, it directly maps to AI developer toolchains, coding agents, and engineering workflows. Key entities and technical topics under continuous tracking include Perplexity, 自蒸馏, 智能体, 后训练.
Source attribution: curated from Perplexity (@perplexity_ai).
