Skip to content

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

Oct 12 · Monday

All papers

Found 206 papers
Filter2

Another 305papers do not have category tags yet and are not included in filtered results.

  1. DefensesarXiv 2610.12602

    Self-Preservation in Language Models

    Proposes Task Bank self-replay to mitigate held-out forgetting in continual instruction tuning

    Published
    Evaluated models
    Qwen3-0.6B, Qwen3-1.7B, Qwen3-4B and others4
    AbstractThe authors claim self-replay can reduce forgetting outside the list, and that on the 4B model new task information helps with target selection, while smaller models do not.

    The authors study how language models, when learning new tasks, can use their own knowledge for broader self-preservation: given an incomplete list and forgetting that extends beyond the list, can the model generate replay and help choose targets.

    Paper analysisRead analysis
  2. DefensesarXiv 2610.12432

    FAITH: A feasibility-aware safety filtering reinforcement learning framework for high-dimensional systems

    Proposes the FAITH safe filtering reinforcement learning framework, training high-dimensional systems to optimize task reward under constraints

    Published
    AbstractFAITH separates safety filtering from reward optimization.

    FAITH is a model-free safe filtering reinforcement learning framework for high-dimensional systems: safety correction and task reward reside in separate modules.

    Paper analysisRead analysis
  3. DefensesarXiv 2610.11352

    RL-ARC:用推理引导的不确定性校准大型推理模型

    提出 RL-ARC,用推理置信度在强化学习中校准推理模型

    Published
    Evaluated models
    Qwen2.5-7B, Qwen3-8B, Llama-8B distilled from DeepSeek-R1
    AbstractRL-ARC 用推理置信度调节答案置信度,作者称可少牺牲精度并改善 ID 与 OOD 校准。

    RL-ARC 是校准感知的强化学习框架,用推理过程的口头置信度调节答案置信度。

    Paper analysisRead analysis
  4. DefensesarXiv 2610.11281

    包络采样:用少量真值标注重校准 LLM 评委以缓解奖励作弊

    提出包络采样,用少量真值标注重标定 LLM 评委以缓解奖励作弊

    Published
    Evaluated models
    Qwen3-4B, Qwen3.8-27B, Qwen2.5-14B-Instruct
    Abstract包络采样把真值标注打到代理尾部,同样预算下缓解 reward hacking。

    作者研究后训练前如何选择真值标注的查询分布,使重标定后的廉价代理更适合被优化。真奖励昂贵时,实践用 LLM judge 代替人类。若失准只发生在基座很少生成的输出上,on-policy 标注修不到这些点,随后的 KL 正则优化会把质量集中到被高估的输出。

    Paper analysisRead analysis
  5. DefensesarXiv 2610.11012

    Redwood Research 提出蒸馏双重困境:用蒸馏让模型坦白隐藏行为或只继承能力

    提出 DFI 与 DFC 蒸馏,定罪隐瞒失配的教师或安全提取能力

    Published
    Evaluated models
    Llama-3.3-70B-Instruct (AuditBench SFT/KTO organisms), Llama-3.3-70B-Instruct (DFI distilled student), Llama-3.1-8B-Instruct (DFI distilled student) and others4

    Abstract概括了论文针对隐瞒失配模型提出的 DFI 与 DFC 机制、主实验发现及适用边界。

    Paper analysisRead analysis
  6. DefensesarXiv 2610.11005

    研究:叙事包装可跨语言绕过 LLM 拒答

    提出 GUISE 基准与 AXIS 训练,防御叙事包装越狱

    Published
    Evaluated models
    Qwen3-1.7B, Qwen3-4B, GLM-4-9B
    AbstractGUISE 分开测包装与语域。

    作者测量叙事包装下的拒答失效,并训练一个在未见包装上拒绝、同时仍回答同风格良性请求的适配器。问题是:安全对齐模型拒绝直说的有害请求,却回答嵌进角色或叙事里的同一请求。文言文被用来把语域和包装拆开。

    Paper analysisRead analysis
  7. DefensesarXiv 2610.10833

    研究:Agent 准时停机不等于有效使用时间预算

    用时钟 harness 和预算塑形 GRPO 改善 Agent 的按时返回

    Published
    Setting
    AI Agent
    Evaluated models
    Qwen3.6-27B, Qwen3-4B, Qwen3.5-9B and others4
    Abstract作者比较提示词、harness 与 GRPO:按时性可改善,但额外时间仍未变成更好解。

    作者研究小规模 LLM 智能体在显式 wall-clock 预算下能否按时返回,并把更多时间换成更好的解。

    Paper analysisRead analysis
  8. DefensesarXiv 2610.10345

    SLDR:通过选择性层恢复与动态路由防御恶意微调

    提出 SLDR,用选择性层恢复与动态路由防御恶意微调

    Published
    Attacker
    Black-box
    Evaluated models
    Llama3.1-8B-Instruct, Llama3-8B-Instruct, Qwen2.5-7B-Instruct and others4
    Abstract作者称 SLDR 能压低有害输出并保持下游效用。

    SLDR 是针对恶意微调的微调后防御:在已对齐模型上做有符号层探测,只在敏感分数最大与最小的两层训练 LoRA,再用 lmax 的最后 token 表示决定是否启用适配器。

    Paper analysisRead analysis
  9. DefensesarXiv 2610.09941

    OrthoPurify 用一步正交投影净化被后门污染的视觉语言模型

    提出 OrthoPurify,用正交投影清除视觉语言模型适配器中的劫持方向

    Published
    Evaluated models
    LLaVA-1.5-7B, Qwen3-VL-8B, LLaVA-1.5-13B and others5
    Abstract一次投影去掉劫持方向。

    OrthoPurify 是部署前的 LVLM 后门净化方法,用一次正交投影去掉被劫持的权重更新方向。

    Paper analysisRead analysis
  10. DefensesarXiv 2610.09793

    用时序逻辑监控工具 Agent

    提出 MFOTL 离线时序验证以监控工具 Agent 的危险动作链

    Published
    Setting
    AI Agent
    Attacker
    Black-box
    Abstract形式化验证可有效识别多步攻击,但其有效性高度依赖轨迹的可信历史记录。

    这篇论文对工具调用大语言模型智能体的安全监控进行了形式化时序验证研究。

    Paper analysisRead analysis
  11. DefensesarXiv 2610.09600

    SafeEvo:从拒答回路视角解释大模型安全对齐机制与演化

    用 SafeEvo 提取拒绝电路,并提出 SCA 仅微调该电路

    Published
    Evaluated models
    Llama-3-8B, Qwen-2.5-7B, Mistral-7B-v0.1 and others4

    AbstractSafeEvo 识别出基座模型中存在的弱安全电路及其演化动态,据此提出的 SCA 算法实现了安全对齐、能力保留与低过度拒绝的平衡。

    Paper analysisRead analysis
  12. DefensesarXiv 2610.09496

    SOUL:用稀疏特征策略遗忘缓解 VLA 模型的状态幻觉

    提出 SOUL 稀疏特征策略遗忘,缓解 VLA 状态幻觉

    Published
    Evaluated models
    OpenVLA, π0.5
    AbstractSOUL 以稀疏特征遗忘 VLA 状态幻觉。

    SOUL用稀疏特征引导的策略 unlearning 处理 VLA 的 state hallucination。

    Paper analysisRead analysis