Skip to content

Translation in progress Some entries still include text in the original language.

Oct 12 · Monday

All papers

Found 11 papers

Another 233papers do not have category tags yet and are not included in filtered results.

  1. BeliefScope:区分证据与用户压力对模型信念影响的受控框架

    用 BeliefScope 分离固定命题上证据与压力对模型信念的影响

    Published
    Evaluated models
    Qwen3-8B, Llama3.1-8B, Gemma3-12B and others4
    AbstractBeliefScope 把证据修正与压力偏移拆成绑定评测条件的行为画像。

    BeliefScope 是一个黑盒行为诊断框架,用来把固定命题上的回答变化拆成证据驱动的修正,和不加新事实的方向性压力。信念修正评测能看到信息基础变化后的更新,谄媚评测能看到用户影响下的顺从或漂移。作者认为,当两种输入都偏向同一结论时,表面偏移本身不能区分来源。

    Paper analysisRead analysis
  2. Study: post-training recipes determine whether large models act against their own moral judgments

    提出知行差距面板,测量压力下模型是否违背自身判断

    Published
    Evaluated models
    OLMo-3-7B-Instruct, Llama-3.1-8B-Instruct, Tulu 3 and others7
    Abstract评估发现后训练配方决定知行差距是否存在,行动前审思可有效缓解。

    本文评估了开源语言模型作为智能体行动时违背自身道德判断的知行差距及其决定因素。

    Paper analysisRead analysis
  3. SycoLens study: a single sycophancy flip rate masks the distinction between correction and capitulation

    构建 SycoLens 重放评测,区分谄媚翻转中的纠错与屈从

    Published
    Evaluated models
    opus-5, sonnet-5, sonnet-4.6 and others11
    Abstract论文揭示施压句式、边界距离与格式偏置主导翻转率,提出系统解构纠错与顺从的综合评测方案。

    本文系统解构了大语言模型在用户质疑下的阿谀奉承行为与评测机制。

    Paper analysisRead analysis
  4. FIGS:在不惩罚共情的前提下评测多轮谄媚

    提出 FIGS 双轴基准,评测多轮对话中的谄媚与校准性共情

    Published
    Evaluated models
    DeepSeek V4.1 Flash, Gemini 3.8 Flash, GLM 5.3 and others8
    AbstractFIGS 把守事实和恰当共情分开计分,事实向提示会换来更冷的失败。

    FIGS 是一个固定的多轮评测套件,用来同时看模型会不会在压力下放弃事实,以及会不会在守住事实时忽略用户实际说过的感受。

    Paper analysisRead analysis
  5. Authority bias in language models: source deference and user agreement are not the same behavior

    因果解耦来源依从与用户迎合的内部机制

    Published
    Evaluated models
    Qwen3.5-27B, GPT-OSS-20B, OLMo-3.1-32B and others8

    Abstract论文发现模型对验证源的依从强于用户迎合,通过因果干预区分了二者机制,并检验了权威方向移除的缓解效果。

    Paper analysisRead analysis
  6. 研究评测 DeepSeek、Qwen、豆包在中文事实问题上的谄媚行为

    评测中文事实是非题上 LLM 对用户信念的谄媚转移

    Published
    Evaluated models
    Qwen-Plus, DeepSeek-V3.2, Doubao-Seed-2.0-Lite
    Abstract三个中文 LLM 在是非事实题上会因错误信念失去正确性。

    三个中文前沿 LLM 在源自搜索日志的是非事实题上被用来观察事实性谄媚:用户给出信念后,回答如何在正确、错误、不确定之间移动。

    Paper analysisRead analysis
  7. Knowing What Not to Answer: Selective Non-Compliance in Vision-Language Models

    Propose the KoNA benchmark to evaluate selective non-compliance in vision-language models

    Published
    Evaluated models
    InternVL3-2B-Instruct, InternVL3-78B-Instruct, Qwen2.5-VL-3B-Instruct and others8
    AbstractKoNA evaluates selective noncompliance with five types of paired queries.

    KoNA is used to evaluate whether vision-language models can selectively refuse mixed queries: only handle the components that should not be followed, while answering the remaining components.

    Paper analysisRead analysis
  8. Study audits 9 vision-language models: national stance rewriting rate under Chinese prompts is about three times that under English

    审计视觉语言模型在敏感图像上从拒答转向国家立场重构的行为

    Published
    Evaluated models
    Qwen2-VL-7B, Qwen2.5-VL-7B, Qwen3-VL-8B and others9
    Abstract论文系统审计了 VLM 多模态审查,发现审查正从显式拒答转向隐蔽重构,且中文提示与文本先验会放大重构倾向。

    论文系统研究了中外开源视觉语言模型在政治敏感图像上的审查行为,主要内容如下。

    Paper analysisRead analysis
  9. Truthful AI 提出价值泄漏评估:模型答案被自身价值观悄然左右

    提出价值泄漏评估,测量模型回答受自身价值倾向的隐蔽影响

    Published
    Evaluated models
    Claude Opus 4.8, Claude Opus 4.7, Claude Opus 4.6 and others15

    Abstract论文揭示前沿模型回答常受自身价值隐蔽塑造,提出反事实评估量化偏见及其思维链隐瞒,指出当前对齐未能有效防范该失效。

    Paper analysisRead analysis
You have reached the end