Skip to content

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

Oct 12 · Monday

All papers

Found 468 papers
Filter2

Attack type

Jailbreaks9Prompt injection73Poisoning and backdoors78Model tampering5Extraction and theft3Detection evasion51Malicious use7

Attack technique

Gradient and representation optimization35Automated search22Encoding and obfuscation12Role-play and persuasion11Conversation structure manipulation4Reasoning-chain manipulation3Cross-modal payloads18Defense-targeted18Dormant triggers27Selection manipulation5

Model risks

Deception and scheming50Reward hacking57Misalignment and value divergence23Dangerous agent actions12Inappropriate refusal3Hallucination and fabrication31Bias and discrimination7

Defense type

Model hardening43Inference-time intervention13Detection and filtering38Instruction–data separation10Permissions and approval10Monitoring and auditing30Watermarking and provenance9

Impact

Harmful content321Unauthorized actions220Data and privacy leaks136Output manipulation344Availability and reliability91Oversight failure124Increased misuse capability19

System type

Models and APIs235LLM applications78AI Agent160coding agent33science agent11web agent6Browser and computer use3Multi-agent systems35Embodied systems and robotics7

Attacker access

Black-box20Gray-box5White-box12

Attack surface / entry point

User input69Web pages and interfaces17Documents and messages31Tool responses20Tool and skill descriptions6Code repositories and dependencies10Retrieval corpus28Memory writes8Inter-agent messages12Images, audio and video51Training data34Weights and gradients21Distribution channels17Inference services and hardware9

Layer

Prompt layer64Model layer118Application layer205Training and data layer76Infrastructure layer5

System component

MCP and skills7Memory25RAG30Guardrails and judges60Monitors43Reasoning chains31Permissions and sandboxing15Fine-tuning and open weights18Vision54Audio10Video7Image generation10Embeddings6Multimodal76
Select any value within a category; different categories must all match. Counts include the other filters you have selected.Another 233papers have no tags yet and are not included in filtered results.
Impact：Output manipulationImpact：Oversight failureClear all

Another 233papers 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
  2. OtherarXiv 2610.13810

    Using subsequence association filtering and breadth-aware selection to suppress hallucinated words in LLM query expansion

    Proposes SAFE-QE, which filters unsupported hallucinated words in LLM query expansion

    arXiv
    2610.13810
    Published
    Layer
    Application layer
    Setting
    LLM applications
    Evaluated models
    BM25, SPLADE, Contriever

    AbstractSAFE-QE suppresses expansion hallucinations with reproducibility and coverage selection; the authors say ranking quality shows no visible cost.

    Paper analysisRead analysis
  4. Analysis and theoryarXiv 2610.13051

    Space beats statistics, retraining beats ranking: auditing training labels under label-only backdoors

    Auditing training labels for label-only backdoors, comparing spaces and statistics, and judging by retraining

    arXiv
    2610.13051
    Published
    Layer
    Training and data layer
    Setting
    Models and APIs
    Evaluated models
    ResNet-32
    AbstractLabel inspection must be done in a space that has not been fitted to these labels, and judged by retrained models rather than detection scores.

    The authors ask: can a defender without trusted data decide which training labels to keep, and how much clean data this destroys, and what the final model accuracy is.

    Paper analysisRead analysis
  5. Risky behaviorarXiv 2610.12822

    Reward hacking as a transmissible strategy in multi-agent coding systems

    Show that reward hacking can propagate through code, memory, and plans in multi-agent coding systems

    arXiv
    2610.12822
    Published
    Layer
    Application layer
    Setting
    Multi-agent systems, coding agent
    Evaluated models
    Qwen2.5-Coder-14B, Gemma-3-12B, Qwen2.5-Coder-7B and others16
    AbstractThe authors state that reward hacking can spread across models and affect routing and aggregation, and that review can reduce delivery.

    Reward hacking induced by coding RL can leave the source agent and enter subsequent collaboration and system coordination.

    Paper analysisRead analysis
  6. AttacksarXiv 2610.12644

    Cancerous backdoor attack: progressive activation via invisible tokens and a stage controller.

    Proposes the Cancerous backdoor, which progressively corrupts reasoning using invisible stage tokens.

    arXiv
    2610.12644
    Published
    Layer
    Training and data layer
    Setting
    Models and APIs
    Evaluated models
    Llama-3.1-8B-Instruct, Qwen2.5-7B-Instruct, Mistral-7B-Instruct-v0.3
    AbstractHidden tokens plus a stage controller make the backdoor progressively corrupt reasoning upon triggering, and the authors believe that looking only at the answer will miss it.

    Cancerous Backdoor Attack changes LLM backdoors from one-time activation to multi-stage reasoning corruption that accumulates progressively with triggers.

    Paper analysisRead analysis
  8. 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
  9. Evaluation and benchmarksarXiv 2610.12360

    The paper proposes the ISE framework to evaluate LLM agents' epistemic humility under knowledge conflicts

    Proposes the ISE framework to evaluate the epistemic humility of LLM agents under knowledge conflicts

    arXiv
    2610.12360
    Published
    Layer
    Application layer
    Setting
    AI Agent
    Evaluated models
    Nemotron-Orchestrator-8B, Claude Sonnet 4.6, GPT-5 and others5
    AbstractISE trajectory evaluations show that high task accuracy across four agents does not correspond to higher cognitive humility.

    What the authors want to test is not whether the agent's final answer is correct, but whether, when a knowledge conflict arises, it recognizes the gap in its trajectory, performs bounded tool follow-up, and writes the unresolved uncertainty to the user.

    Paper analysisRead analysis
  10. Analysis and theoryarXiv 2610.12313

    The study tests the sensitivity of five large-model compliance systems to perturbations of regulatory rules.

    Diagnose whether compliance system rulings change with perturbations to regulatory rules

    arXiv
    2610.12313
    Published
    Layer
    Prompt layer
    Setting
    Models and APIs
    Evaluated models
    Llama-3.3-70B-Instruct, Llama-3.1-8B-Instruct, Qwen2.5-7B-Instruct and others5
    AbstractAcross five models, many correct verdicts are largely unchanged by deleting or replacing negation rules.

    Use behavioral counterfactuals and activation probes to audit whether compliance rulings rely on the given rules. The authors aim to separate two kinds of seemingly identical correct rulings: those that use the given rules, and those obtained from case surface features or training patterns.

    Paper analysisRead analysis
  11. Analysis and theoryarXiv 2610.12125

    The researchers found that optimizing the teaching fit metric makes LLM tutors repeat the same option, with expert scores dropping from 4.46 to 3.03

    Analyzing why optimizing teaching adaptation metrics collapses tutoring choices and diverges from expert judgment

    arXiv
    2610.12125
    Published
    Layer
    Training and data layer
    Setting
    Models and APIs
    Evaluated models
    Gemini 3.1 Pro Preview, Qwen3-32B
    AbstractMetrics rise while expert scores fall: the optimum of the separable average is repetition, and the correction mainly rewrites the output projection.

    After turning teaching adaptation metrics into a fine-tuning objective, the metrics rise while scores from trained educators decline.

    Paper analysisRead analysis
  12. 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
  15. AttacksarXiv 2610.11932

    A Fudan team measures the citation fragility of AI search: ordinary posts can enter citations and answer text

    Measures citation fragility in AI search, demonstrating that ordinary public posts can enter citations and answers.

    arXiv
    2610.11932
    Published
    Layer
    Application layer
    Setting
    LLM applications
    Evaluated models
    Grok, Doubao, ChatGPT and others10

    AbstractThe authors claim the paper quantifies vulnerabilities in AI search caused by citation preferences and low posting thresholds; tests show public web posts can penetrate AI citations, and they call for auditing the citation supply chain.

    Paper analysisRead analysis
  16. 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
  17. DefensesarXiv 2610.11754

    仅凭输出审查无法验证什么：面向研究智能体的研究契约

    提出研究合同，把已批准选择绑定到执行证据并由检查器核验

    arXiv
    2610.11754
    Published
    Layer
    Application layer
    Setting
    science agent
    Evaluated models
    OpenAI gpt-5.6-sol, OpenAI gpt-5.6-terra, OpenAI gpt-5.6-luna and others18
    Abstract研究合同暴露四类输出单独建不立的关系。

    研究合同用来核验计算科学研究智能体是否仍在已批准研究之内，不判断科学是否为真。。

    Paper analysisRead analysis
  18. DefensesarXiv 2610.11634

    LTBD：用可学习信任边界分隔符防御提示注入

    提出 LTBD，用可学习分隔符隔离可信指令与不可信数据

    arXiv
    2610.11634
    Published
    Layer
    Application layer
    Setting
    LLM applications
    Evaluated models
    Llama3-8B-Instruct, Llama3.1-8B-Instruct, Falcon3-7B-Instruct and others4
    AbstractLTBD 只训练四个分隔符嵌入。

    LTBD 是一种不改基座权重的提示注入防御：只学习用来标出可信指令与不可信数据边界的四个 token 嵌入。

    Paper analysisRead analysis
  19. Evaluation and benchmarksarXiv 2610.11514

    SSCBench：评估工具调用 LLM Agent 故障注入测试证据有效性的测量协议与基准

    提出 SSCBench，评测工具调用 Agent 故障注入测试的证据有效性

    arXiv
    2610.11514
    Published
    Layer
    Application layer
    Setting
    AI Agent
    Evaluated models
    deepseek-flash, deepseek-v4-pro, gpt-5.6-luna and others5
    Abstract解释采纳前应报告该次执行的证据条件，以及支撑该主张的观察总体。

    SSCBench 研究工具使用智能体的故障注入结果，在什么证据条件下可以被解释。

    Paper analysisRead analysis
  20. DefensesarXiv 2610.11464

    研究者提出可检查评分器与自改进 Agent 共同演化方法

    演化可检查评分器并与自改进 Agent 共同演化，抑制奖励作弊

    arXiv
    2610.11464
    Published
    Layer
    Application layer
    Setting
    AI Agent
    Evaluated models
    Claude Opus 4.7
    Abstract十项锚点演化可检查评分器。

    开放任务没有验证器时，被演化的是评分器，而不是再补一条手写 rubric。

    Paper analysisRead analysis
Load more