Skip to content

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

Oct 12 · Monday

All papers

Found 3 papers
Filter5

Model risks

Deception and scheming5Reward hacking3Misalignment and value divergence12Dangerous agent actions3Inappropriate refusal3Hallucination and fabrication19Bias and discrimination6

Impact

Harmful content47Output manipulation3Availability and reliability8

System type

Models and APIs3coding agent0web agent0
Select any value within a category; different categories must all match. Counts include the other filters you have selected.Another 232papers have no tags yet and are not included in filtered results.
Model risks：Inappropriate refusalImpact：Output manipulationSystem type：Models and APIsSystem type：coding agentSystem type：web agentClear all

Another 232papers do not have category tags yet and are not included in filtered results.

  1. DefensesarXiv 2609.33086

    研究者提出基于评分量表的可解释奖励框架，将宪法转化为可调对齐目标

    提出可解释 rubric 奖励，将宪法转为可调权重并用于监督微调

    arXiv
    2609.33086
    Published
    Layer
    Training and data layer
    Setting
    Models and APIs
    Evaluated models
    Gemma 3 12B, Qwen3 14B, GPT-OSS 20B
    Abstract宪法级可解释奖励让优先级可检查、可改写，并带入训练后的行为。

    作者给出一条从通用宪法到可调训练奖励的路径，用来检查并改写模型实际被优化的优先级。

    Paper analysisRead analysis
  2. Evaluation and benchmarksarXiv 2609.04720

    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

    arXiv
    2609.04720
    Published
    Layer
    Model layer
    Setting
    Models and APIs
    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
  3. Evaluation and benchmarksarXiv 2608.11816

    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

    arXiv
    2608.11816
    Published
    Layer
    Model layer
    Setting
    Models and APIs
    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
You have reached the end