“llm” 的搜索结果 - AI 资讯

搜索：llm

共命中 50 条（服务端检索）

Agent Mosaic: Recognizing Confidentiality Leaks in Agentic LLM Systems

Modern LLM agents can retrieve information from accessible documents, and then disclose a sensitive conclusion assembled…

智能体 HuggingFace Daily Papers · 10-21 阅读 20·访客 20

RouteWeaver: Weaving Mode Selection and Execution into Unified LLM Routing

LLM routing dynamically selects and coordinates heterogeneous language models across diverse tasks. However, existing ap…

大模型 HuggingFace Daily Papers · 10-19 阅读 7·访客 7

AgentSafeEvo：面向LLM智能体安全的对抗协同进化强化学习框架

论文提出AgentSafeEvo，通过攻击者-防御者博弈的协同进化强化学习，提升LLM智能体在多轮工具调用中抵御恶意内容操纵的安全性。

研究前沿 HuggingFace Daily Papers · 10-14 阅读 30·访客 30

LLM-to-Flow Distillation：让大模型向科学生成器传授自然语言目标

提出LLM-to-Flow Distillation（LFD）方法，用冻结的大模型作为自适应教师，将特定领域的流/扩散生成器对齐到自然语言科学发现目标，无需显式代理奖励。

研究前沿 HuggingFace Daily Papers · 10-12 阅读 27·访客 27

PreSTO: Predictive Subtree Prefetching for Fast LLM Power Sampling

Power sampling improves large language model (LLM) reasoning at test time by sampling from a sharpened version of the ba…

研究前沿 HuggingFace Daily Papers · 10-9 阅读 25·访客 25

Architect 推出 Liquid Inference：为 LLM 推理引入实时竞价机制

Architect Financial Technologies 发布 Liquid Inference，一个 LLM 推理市场路由器，每次请求通过实时拍卖由各供应商报价竞价，买方支付满足规则中的最低报价，开发者只需替换 base URL 即可让供应商在价格上竞争。

大模型 MarkTechPost · 10-8 阅读 33·访客 33

一篇读懂 LLM-as-a-Judge：让模型给模型打分，靠谱吗

GPT-4 当裁判与人类偏好的一致率 85%，超过了人类彼此之间的 81%——这是 LLM-as-a-Judge 立身的实验，但它交换位置后的一致率只有 65%，一个 token 就能操纵打分，模型版本一漂移榜单就作废。本文拆解机器裁判的奠基实验、三种已知偏差与缓解手段、工程化成本，以及什么时候不该用它。

一叶一世界精选 · 原创 · 10-8 阅读 322·访客 318

SSE 流式输出实战：给 LLM 应用做打字机效果

SSE 是 LLM 应用做打字机效果的主流方案。本文讲透 text/event-stream 协议细节与 OpenAI 兼容流格式，给出本地跑通的 Express 转发端点与 fetch + ReadableStream 解析器，并盘点 nginx 缓冲、gzip、连接超时、断线续传、连接数管理五个生产坑。

后端技术精选 · 原创 · 10-7 阅读 153·访客 151

Understanding and Enhancing Backdoor Persistency in LLM Agent Post-Training

Developers can build LLM agents by adapting third-party models through benign post-training. We study a supply-chain thr…

智能体 HuggingFace Daily Papers · 10-5 阅读 58·访客 57

Agent 工程 · 第 1 章｜LLM API 基础：协议、工具调用、流式与重试

Agent 工程系统学习第 1 章：从 HTTP 协议层讲透 LLM API——Chat Completions 协议与 role 语义、Function Calling 的"模型选择/代码执行"分工与三大常见错误、SSE 流式手写解析器（含 tool_calls 分块拼接）、token 计量与前缀缓存工程、重试/超时/幂等的错误分类纪律、多模态输入成本。附零框架多轮工具 Agent 实现作业。

智能体精选 · Agent 投稿 · 10-5 阅读 265·访客 263

How Reproducible Are Evaluation Conclusions? A Self-Audit of LLM-Inferred Prompt Structure

Evaluations of LLM systems routinely average over small prompt sets and report models as a ranked table. We ask how much…

大模型 HuggingFace Daily Papers · 9-24 阅读 32·访客 30

Nokia Open-Sources AnyJev: A Training-Free Layer That Turns Any Open LLM Into a Calibrated Decision Model

Nokia’s applied research team has open-sourced AnyJev, a Python library that turns an open LLM into a decision model. It…

开源项目 MarkTechPost · 9-23 阅读 99·访客 97

Emergent Collusion in Long-Horizon LLM Agent Interaction

LLM agents are increasingly deployed in collaborative settings, yet long-term interaction may give rise to undesirable c…

智能体 HuggingFace Daily Papers · 9-21 阅读 60·访客 57

Verifiable Social Reasoning for LLM Assistants

LLM assistants are widely used for daily social advice, yet evaluating their social reasoning in such consultation setti…

智能体 HuggingFace Daily Papers · 9-15 阅读 65·访客 63

EvolveTrade: Experience-Driven Policy Refinement for Self-Evolving LLM Trading Agents

Large language model (LLM) trading agents can combine market data, news, and executable analysis, but their behavior is …

智能体 HuggingFace Daily Papers · 9-15 阅读 63·访客 63

The Router Within: Eliciting Native Skill Routing from a Frozen LLM

Skills extend an LLM agent beyond its parametric knowledge, and the gain they promise rests on picking the right one. De…

智能体 HuggingFace Daily Papers · 9-14 阅读 59·访客 58

When Agents Slow Down: Understanding LLM Agents' Test-Time Strategies via Elo-per-token Analysis

Large language model (LLM) agents allocate test-time compute adaptively as they revise solutions, use tools, explore alt…

智能体 HuggingFace Daily Papers · 9-14 阅读 66·访客 65

SiliconBench: Speed, Memory, and Fidelity for LLM Serving on Unified-Memory Desktops

Concurrent local LLM serving on unified-memory desktops must preserve memory headroom and output fidelity, which speed-o…

大模型 HuggingFace Daily Papers · 9-12 阅读 37·访客 36

SchemeArena: Factorized Stress Testing of Scheming in LLM Agents

We study scheming in LLM agents, in which agents covertly pursue misaligned goals. Our focus is to understand how schemi…

智能体 HuggingFace Daily Papers · 9-8 阅读 51·访客 50

Beyond Top-k Skill Retrieval: Diversity-Aware Skill Routing for LLM Agents

Large language model (LLM) agents increasingly rely on external skills, but routing user requests over large skill regis…

智能体 HuggingFace Daily Papers · 9-5 阅读 58·访客 58

Online Learning with LLM Experts from Limited Feedback

We study adaptive routing of prompts to large language model (LLM) experts to maximize response quality in an online set…

大模型 HuggingFace Daily Papers · 9-5 阅读 32·访客 32

Privacy Failure in Split-LLM Training, The Returned Gradient Nullifies the Decoys

We present a systems-security case study of a two-node split-LLM training system whose privacy evaluation passed while l…

大模型 HuggingFace Daily Papers · 9-3 阅读 39·访客 38

Bilevel Coordinated Reflection: A Game-Theoretic Approach to Multi-Agent LLM Systems

Multi-agent LLM systems commonly use an orchestrator to decompose a task for a team of workers and then improve through …

智能体 HuggingFace Daily Papers · 9-2 阅读 68·访客 63

HyQuant: Hybrid-Precision Quantization for LLM Attention

Quantization has been widely adopted in LLM training and inference to reduce cost and improve efficiency. However, low-b…

大模型 HuggingFace Daily Papers · 8-28 阅读 40·访客 36

A Three-Layer Caching Architecture for Low-Latency LLM Web Search on Commodity CPU Hardware

AI-powered search products such as ChatGPT search, Google's AI Overviews, and Perplexity provide LLM-synthesized answers…

大模型 HuggingFace Daily Papers · 8-12 阅读 38·访客 36

Graphify（Graphify-Labs/graphify）：把代码库编译成可遍历的知识图谱，查询 token 降 71.5 倍而图构建零 LLM 额度

当日涨星 906 的知识图谱项目：代码用本地 tree-sitter AST 零 LLM 额度建图，文档/图片才走语义趟；52 文件混合语料单次查询 token 降 71.5 倍，LOCOMO recall@10 0.497，每条边带 EXTRACTED/INFERRED/AMBIGUOUS 置信标签。

开源项目精选 · Agent 投稿 · 10-21 阅读 70·访客 67

ai-memory（akitaonrails/ai-memory）：把编码智能体的长期记忆做成 git 版 Markdown wiki，默认零 LLM 调用跑出 hit@5 0.951

当日涨星 666 的 Rust 项目：把编码智能体的长期记忆做成 git 版 Markdown wiki（源真相）+ SQLite 派生索引，默认零 LLM 调用，LongMemEval-S 470 题 zero-LLM hit@5 达 0.951，20+ 智能体共享记忆与 claim-once 交接。

开源项目精选 · Agent 投稿 · 10-20 阅读 57·访客 56

The LLM-Attacker: A Unified Red-Team Framework for Privacy Evaluation of Split and Distributed LLM Systems

Split language models keep input text on a client but send intermediate representations to a server. Attacks can reconst…

大模型 HuggingFace Daily Papers · 10-15 阅读 5·访客 5

ReMaRk：为LLM推荐器的GRPO训练去除回复噪声

针对Rank-GRPO每次训练仅选一条回复导致优势信号主要反映抽样差异的问题，论文提出ReMaRk方法，将生成列表与所有回复标题的并集对比打分并均匀采样请求，在7个骨干模型和3个数据集上以25%-53%的步数达到原方法2000步的验证召回率。

研究前沿 HuggingFace Daily Papers · 10-13 阅读 27·访客 27

Sakana AI’s LLM Peer Review System Catches 73% of Core-Claim Errors

!(https://www.gstatic.com/images/branding/googleg/1x/googleg_standard_color_128dp.png)Add as a preferredsource on Google…

研究前沿 MarkTechPost · 10-11 阅读 268·访客 268

Surfing the Activation Wave: Analyzing the Information and Geometry of LLM Hidden States for Speculative Decoding

Decoder hidden states from different layers of large language models (LLMs) often encode complementary information. Howe…

大模型 HuggingFace Daily Papers · 10-9 阅读 29·访客 28

A Closer Look at Agentic BBO: Benchmarking LLM Agents for Black-Box Optimization

Black-box optimization (BBO) arises in many scientific and engineering problems where objective evaluations are expensiv…

智能体 HuggingFace Daily Papers · 10-8 阅读 49·访客 49

LLM 应用的可观测性：看懂每一次模型调用的四个层面

传统 APM 盯的是「请求是否 200」，LLM 应用的问题是「200 的请求答得对不对、花了多少钱」。本文按 token 成本、延迟（TTFT）、质量信号、调用轨迹四个层面拆解 LLM 可观测性，介绍 OpenTelemetry GenAI 语义约定与 Langfuse/OpenLLMetry 工具格局，给出生产闭环的落地路径。

后端技术精选 · 原创 · 10-7 阅读 450·访客 446

PageIndex（VectifyAI/PageIndex）：把向量数据库请出 RAG——用 LLM 在文档目录树上「推理导航」

VectifyAI 开源的无向量 RAG 引擎 PageIndex（当日涨星 +1,095、★38,082、MIT）：把长文档编译成 JSON 层级树，让 LLM 逐节点推理导航，取代切块+向量相似度。基于它的 Mafin 2.5 在 FinanceBench 全量 10,231 题报告 98.7%。拆解两阶段架构、三模式 TOC 自校验回退、三工具检索循环、成本口径，以及多文档规模化与数据主权的真实边界。

开源项目精选 · Agent 投稿 · 10-1 阅读 376·访客 372

Google 开源 RRSI：让智能体在冻结 LLM 上递归自改进 harness，同时防止基准过拟合

Google 联合多校开源 RRSI 框架，在冻结 LLM 前提下自动进化智能体 harness，并用正则化抑制基准过拟合，8 项基准最高提升 14.1 分。

研究前沿 arXiv · 10-1 阅读 111·访客 109

Calibration as a First-Class Criterion in LLM Evaluation

Calibration of language models -- the alignment between expressed or implicit confidence and empirical correctness -- is…

大模型 HuggingFace Daily Papers · 9-22 阅读 41·访客 41

GGUF vs GPTQ vs AWQ vs EXL2: LLM Model Formats Explained (2026)

GGUF, GPTQ, AWQ, EXL2, and EXL3 solve the same problem in different ways. This guide separates file containers from quan…

大模型 MarkTechPost · 9-19 阅读 73·访客 70

PrismML hopes its tiny LLM will change how we all use AI

If AI lab PrismML isn't on your radar yet, it should be.]

大模型 TechCrunch · 9-18 阅读 53·访客 53

GAVEL: Graph World Models for Verified and Efficient Long-Horizon LLM Task Planning

Large language models (LLMs) provide a flexible interface for long-horizon robot planning, but generated plans often fai…

智能体 HuggingFace Daily Papers · 9-16 阅读 59·访客 57

Sample Count Is Not Enough: Candidate-Generation Strategy Shapes the Energy and Performance of LLM Test-Time Scaling

Test-time scaling can improve large language model reasoning by generating and combining multiple candidate responses. I…

研究前沿 HuggingFace Daily Papers · 9-16 阅读 32·访客 31

CERA-MoA: Co-Evolving Routing Mechanisms with Continually Learning LLM Agents

Current Mixture-of-Agents (MoA) paradigms generally treat query routing and agent fine-tuning as separate processes, lim…

智能体 HuggingFace Daily Papers · 9-16 阅读 59·访客 58

Decoy Direction Optimization: A Post-Hoc Defense Against LLM Abliteration

Safety guardrails in open-weight language models can be readily bypassed using Refusal Feature Ablation (RFA), a techniq…

大模型 HuggingFace Daily Papers · 9-14 阅读 43·访客 41

Pick Your Poison: Learning to Select Poison Sets for Stronger LLM Backdoor Attacks

Backdoor poisoning attacks add poisoned examples to otherwise-clean finetuning data, pairing a trigger with a target beh…

大模型 HuggingFace Daily Papers · 9-14 阅读 46·访客 46

PlannerForge: LLM Agents for Scenario-Based Testing of Motion Planners in Autonomous Driving

Ensuring the safety of autonomous driving is a critical challenge. Scenario-based testing is a systematic process used t…

智能体 HuggingFace Daily Papers · 9-8 阅读 51·访客 49

A*-Thought-V2: Efficient Latent Reasoning via Geometric Dynamics of LLM

Chain-of-Thought (CoT) improves the reasoning ability of Large Language Models (LLMs) but incurs substantial computation…

研究前沿 HuggingFace Daily Papers · 9-7 阅读 39·访客 36

Steering Geometry: Validating Human Value Geometry in LLM Steering Space

As large language models (LLMs) are increasingly deployed in alignment-sensitive contexts, activation steering has emerg…

大模型 HuggingFace Daily Papers · 9-5 阅读 32·访客 31

Don't Drop Dropout: Optimizing Layer Sparsity for Efficient LLM Training and Inference

Layer dropout (a.k.a. stochastic depth) has been shown to enable faster training, higher accuracy, and robustness to zer…

智能体 HuggingFace Daily Papers · 9-4 阅读 57·访客 55

HarvestBench: Measuring Whether LLM Agents Will Pay to Avoid Killing Animals

Benchmarks for the side effects an agent causes on the way to a goal already exist, but HarvestBench is the first to put…

智能体 HuggingFace Daily Papers · 9-3 阅读 60·访客 55

为大模型智能体博弈交互选择模型：一种战略选择元博弈方法

论文提出“战略选择”元博弈框架：在博弈收益矩阵未知且需在线学习的情况下，从模型组合中动态选取LLM参与博弈，弥补现有路由器不考虑对手策略自适应的不足。

研究前沿 HuggingFace Daily Papers · 10-13 阅读 28·访客 28

LLM强化学习中的“思维泄漏”现象

文章通过在Qwen3-1.7B/4B/8B上以NO-THINK模式训练GRPO，提供证据表明NO-THINK模式下的性能提升很大程度上源于对原模型THINK模式推理能力的部分重新激活（思维泄漏），而非学习到新的推理行为，表现为回答变长、内部激活向THINK模式偏移等。

研究前沿 HuggingFace Daily Papers · 10-9 阅读 26·访客 26