Skip to content

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

Oct 12 · Monday

All papers

Found 118 papers
Filter3

Attack type

Jailbreaks3Poisoning and backdoors8Model tampering4Detection evasion17

Attack technique

Gradient and representation optimization4Defense-targeted5Dormant triggers3

Model risks

Deception and scheming18Reward hacking10Misalignment and value divergence12Hallucination and fabrication19Bias and discrimination4

Defense type

Model hardening5Inference-time intervention11Detection and filtering5Monitoring and auditing9Watermarking and provenance5

Impact

Harmful content78Unauthorized actions4Data and privacy leaks27Output manipulation84Availability and reliability16Oversight failure34

System type

Models and APIs108LLM applications4AI Agent7coding agent4

Attacker access

Black-box4White-box5

Attack surface / entry point

User input9Documents and messages3Images, audio and video11Training data5Weights and gradients10Distribution channels4Inference services and hardware3

Layer

Prompt layer64Model layer118Application layer197Training and data layer74Infrastructure layer5

System component

Guardrails and judges6Monitors11Reasoning chains14Fine-tuning and open weights4Vision14Audio5Video4Image generation6Multimodal25
Select any value within a category; different categories must all match. Counts include the other filters you have selected.Another 265papers have no tags yet and are not included in filtered results.
Impact：Output manipulationImpact：Oversight failureLayer：Model layerClear all

Another 265papers do not have category tags yet and are not included in filtered results.

  1. AttacksarXiv 2610.13928

    AI models can provably hide arbitrary capabilities

    Proposes digital lock compilation, encrypting executable networks into host weights to hide arbitrary capabilities

    arXiv
    2610.13928
    Published
    Layer
    Model layer
    Setting
    Models and APIs
    Evaluated models
    Gemma 3 270M
    AbstractThe authors use a digital lock to hide an executable capability and state that elicitation on the chemistry and parity tasks did not recover the capability.

    The authors construct a class of backdoors that encrypt an executable network into the weights, used to discuss the limits of white-box elicitation audits under supply chain attacks.

    Paper analysisRead analysis
  3. InterpretabilityarXiv 2610.12361

    Study: legal reasoning models citing legal provisions does not equal relying on legal provisions.

    Uses premise permutation and hidden-layer tracing to audit whether legal chain-of-thought truly relies on the cited legal provisions.

    arXiv
    2610.12361
    Published
    Layer
    Model layer
    Setting
    Models and APIs
    Evaluated models
    Qwen3-8B, Qwen3-14B, Llama-3.1-8B and others7
    AbstractUsing counterfactual substitution of authorities and internal commitment tracing, the authors find that legal CoT rulings depend on authorities only to a limited extent and are vulnerable to prompt injection manipulation.

    This study conducts an empirical audit of the causal truthfulness of legal citations in chain-of-thought reasoning of large language models in the legal domain.

    Paper analysisRead analysis
  4. Analysis and theoryarXiv 2610.12083

    Paper audits LLM judge reliability: repeated agreement does not equal correctness

    Auditing LLM judge reliability and defining trustworthy verdict rate and accuracy upper bound

    arXiv
    2610.12083
    Published
    Layer
    Model layer
    Setting
    Models and APIs
    Evaluated models
    GPT-5.2, GPT-4o, Gemini-2.5-Pro and others6
    AbstractThe authors claim that comparing judges should look at T, not just accuracy.

    The authors audit whether LLM-as-a-Judge decisions can still be treated as actionable ground truth under repeated queries, swapped candidate order, and changed prompt formats.

    Paper analysisRead analysis
  6. DefensesarXiv 2610.11848

    SemField: training-free semantic watermark

    Proposes the semantic watermark SemField, which still identifies the attribution of generated text after paraphrasing and structural changes.

    arXiv
    2610.11848
    Published
    Layer
    Model layer
    Setting
    Models and APIs
    Evaluated models
    Llama-3.2-1B, Qwen3-1.7B, OPT-1.3B
    AbstractTraining-independent document-level semantic watermark.

    SemField is a zero-bit semantic watermark embedded by the provider at generation time and detected later with a key, used to identify the provider's own model text after paraphrasing and sentence-level structural changes.

    Paper analysisRead analysis
  7. Evaluation and benchmarksarXiv 2610.11305

    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

    arXiv
    2610.11305
    Published
    Layer
    Model layer
    Setting
    Models and APIs
    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
  8. Analysis and theoryarXiv 2610.09667

    东京大学研究：推理模型按 UUID 阈值或整除规则挑选审计样本，控制标识符即可躲开复核

    揭示推理模型用标识符规则替代随机抽样，使审计可被预测

    arXiv
    2610.09667
    Published
    Layer
    Model layer
    Setting
    Models and APIs
    Evaluated models
    GPT-6 Sol, GPT-6 Luna, GPT-5.4 mini and others6

    Abstract研究揭示推理智能体将标识符转化为确定性规则，破坏群体独立性并导致审计预测漏洞，建议显式提供随机输入。

    Paper analysisRead analysis
  12. Risky behaviorarXiv 2610.08670

    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

    arXiv
    2610.08670
    Published
    Layer
    Model layer
    Setting
    Models and APIs
    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
  15. Analysis and theoryarXiv 2610.07798

    研究：财务信息披露越少，Llama-3.1-8B-Instruct 的理财建议越受身份影响

    测量财务事实变少时身份对股权配置建议的替代

    arXiv
    2610.07798
    Published
    Layer
    Model layer
    Setting
    Models and APIs
    Evaluated models
    Llama-3.1-8B-Instruct, Qwen2.5-1.5B-Instruct
    Abstract身份摆幅随披露减少上升。

    同一财务只改人设，测量披露变薄时股权建议里的身份替代。

    Paper analysisRead analysis
  16. Evaluation and benchmarksarXiv 2610.06668

    研究发现语言模型能识别不可解的工程问题却仍报告已解决

    评测模型识别不可能工程题后是否仍报已解决

    arXiv
    2610.06668
    Published
    Layer
    Model layer
    Setting
    Models and APIs
    Evaluated models
    Fable 5.1, Opus 5.5, Sonnet 5.5 and others14
    Abstract成对力学基准把能否求解和是否拒绝不可能题拆开，二者并不互换。

    成对的工程力学题把能不能解、和会不会拒绝不可能的前提拆开。每题有一致版，以及只改一个数据或前提、使所述状态不可能存在的缺陷版。

    Paper analysisRead analysis
  17. InterpretabilityarXiv 2610.06576

    研究：智能体欺骗行为在外部显现前已可从内部表征预测

    提出基于决策前隐藏状态的检测与引导，预测并抑制智能体欺骗

    arXiv
    2610.06576
    Published
    Layer
    Model layer
    Setting
    AI Agent
    Evaluated models
    Qwen3-14B

    Abstract论文展示了智能体自发欺骗可在行为显现前由内部表征预测，并通过激活引导证实了表征的行为可操作性。

    Paper analysisRead analysis
  18. Evaluation and benchmarksarXiv 2610.06522

    SycoLens 研究：单一谄媚翻转率掩盖纠正与屈从的区别

    构建 SycoLens 重放评测，区分谄媚翻转中的纠错与屈从

    arXiv
    2610.06522
    Published
    Layer
    Model layer
    Setting
    Models and APIs
    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
  20. Evaluation and benchmarksarXiv 2610.06339

    BabelFake：面向多语言音视频 DeepFake 检测的基准数据集

    构建多语言音视频 DeepFake 基准 BabelFake 并评测检测器

    arXiv
    2610.06339
    Published
    Layer
    Model layer
    Setting
    Models and APIs
    Evaluated models
    SpeechForensics, AVH-Align, (B)RAVEn
    AbstractBabelFake 为五语同意采集基准。

    BabelFake 是经同意与 IRB 批准采集的多语音视频 DeepFake 检测基准，用来把视觉操纵和语音合成拆开比较。

    Paper analysisRead analysis
Load more