Skip to content
Hot eventLive

OLMo-Detect:面向LLM成员推断的多阶段受控基准

1 reports1 sources16 hr ago updated

Get the story

AI overview

2026年10月5日,arXiv Computation and Language 发布一手报道,介绍 OLMo-Detect——一个面向大语言模型成员推断的多阶段、混杂因素受控基准。该基准基于完全开放的 OLMo 2 流水线构建,覆盖预训练、中训练与后训练阶段,并在三个关键轴上对齐成员与非成员样本,同时通过 infini-gram 严格过滤非成员,以降低混杂因素对成员推断评估的干扰。目前报道仅披露上述设计与构建信息,未涉及实验结果或后续更新。

Generated from reports · updated 16 hr ago

Timeline

Follow the coverage from different angles.

Oct 5, 2026
  1. arXiv · Computation and Language
    OLMo-Detect:面向大语言模型成员推断的多阶段、混杂因素受控基准

    OLMo-Detect 基于完全开放的 OLMo 2 流水线构建,覆盖预训练、中训练与后训练,并在三个关键轴上对齐成员与非成员样本,同时通过 infini-gram 严格过滤非成员。

Heat trend

Current heat 6·Comparable peak 10(Oct 5)·Comparable change over 24 hours –

02.557.510Oct5Oct5Oct5Oct6

The trend compares only the same participants observed continuously; its range may be smaller than the current heat count. Move or click on the chart to inspect hourly heat; use the left and right arrow keys to switch.