独立 AI 资讯库
从自动抓取数据库实时读取，适合快速浏览、检索和后续扩展成完整资讯站。
arXiv:2608.00419v1 Announce Type: new Abstract: Large language models deployed in real-time, regulated settings face knowledge staleness, catastrophic forgetting, hallucination, and weak feedback loops. We present a unified, pattern-driven LLMOps architecture integrating real-time data ingestion, c
arXiv:2608.00316v1 Announce Type: new Abstract: Bayesian optimization (BO) has become the standard tool for sample-efficient optimization and owes its efficiency to uncertainty-aware search driven by generic statistical priors. Richer domain priors can improve BO in principle, but encoding them thr
arXiv:2608.00301v1 Announce Type: new Abstract: Error-penalized scoring rules ($+1$ for a correct answer, $-\lambda$ for a wrong one, $0$ for abstaining) are increasingly prescribed against hallucination: a rational agent facing such a rule answers exactly when its correctness probability exceeds C
arXiv:2608.00198v1 Announce Type: new Abstract: Environmental time-series causal discovery requires expert decisions about method choice, conditional-independence tests, lag horizons, sample-size adequacy, multiple-testing control, and evidence interpretation. Applied inconsistently across datasets
arXiv:2608.00175v1 Announce Type: new Abstract: Deep reinforcement learning (RL) agents achieve strong performance by optimizing scalar reward functions. However, once deployed, the policies of these RL agents are often rigid and costly to adapt to new performance criteria. For instance, an agent t
arXiv:2608.00152v1 Announce Type: new Abstract: Predicting the magnitude of a CRISPRi perturbation's transcriptomic effect on held-out target genes is an important open problem in single-cell biology. Recent work has documented that simple baselines often match or exceed deep perturbation predictor
arXiv:2608.00107v1 Announce Type: new Abstract: Agentic systems must repeatedly decide whether to answer directly, decompose a task, invoke a tool, execute code, delegate to a specialist, verify an intermediate result, or recover from failure. These meta-decisions affect not only task success but a
arXiv:2608.00106v1 Announce Type: new Abstract: Agentic systems must decide not only what answer to produce, but which reasoning and execution operations should precede it. A controller may answer directly, decompose a request, retrieve evidence, execute code, delegate to a specialist, or verify an
arXiv:2608.00485v1 Announce Type: new Abstract: Recent Text-to-SQL systems increasingly rely on multi-turn interaction, execution feedback, and reinforcement learning. However, most existing methods use execution correctness only as a trajectory-level reward, which provides limited guidance for ide
arXiv:2608.00432v1 Announce Type: new Abstract: Deep research agents are often trained on expensive, environment-grounded tool-use trajectories that require repeated retrieval, document inspection, and report evaluation. We introduce Deep Research Pretraining (DRP), an offline framework that derive
arXiv:2608.00355v1 Announce Type: new Abstract: Progress in large language models is often summarized using a single scalar measure, such as a time horizon, a latent ability estimate, or an aggregate benchmark score. These summaries capture the overall performance, but they do not test whether prog
arXiv:2608.00218v1 Announce Type: new Abstract: Agentic LLMs exhibit three consequential tool-use failures: invalid arguments (validity), unnecessary calls (over-calling), and omitted calls when tools are needed (missing). We find that a small, failure-specific set of MLP neurons could distinguish
arXiv:2608.00036v1 Announce Type: new Abstract: Real-world document tasks often ask professionals to answer questions from annual reports, regulations, clinical guidelines, and technical manuals that span hundreds or thousands of pages. Some questions also require comparing related reports. Reliabl
arXiv:2608.00030v1 Announce Type: new Abstract: Specialised retrieval agents typically surface higher quality results than general-purpose search, but selecting the optimal agent for a given query remains an open problem. Current approaches route queries based on inferred topic or intent, however i
arXiv:2608.00023v1 Announce Type: new Abstract: Social simulations built from language-model agents need role-conditioned behavior that can be checked before agents are placed into a simulated population. We introduce an activation-steering screening workflow for role-conditioned agents: define a r
arXiv:2608.00013v1 Announce Type: new Abstract: Choosing the right large language model (LLM) backbone is the most consequential decision when building a vision-language model (VLM), yet it remains fundamentally unprincipled: compute-based scaling laws fail to generalize across model families, and
arXiv:2608.00012v1 Announce Type: new Abstract: Multimodal Large Language Models (MLLMs) are increasingly used to interpret Earth observation data, yet their capability to support real-world disaster emergency response remains insufficiently evaluated. Existing remote sensing benchmarks largely rel
arXiv:2608.00009v1 Announce Type: new Abstract: Long-term memory remains a critical bottleneck for conversational AI agents, whose finite context windows cannot support coherent recall across thousands of turns. We present AgentMemBench, a unified, reproducible benchmark evaluating five memory mana
arXiv:2608.00007v1 Announce Type: new Abstract: Equipping Large Language Models (LLMs) with human-like personas is crucial for agentic applications, such as role-play and user simulation. Traditional prompt-based methods rely on descriptive conditioning by injecting static textual profiles, which o
arXiv:2608.00005v1 Announce Type: new Abstract: Peer review at major venues is under unprecedented submission pressure, motivating the use of large language models (LLMs) as review assistants. Existing LLM-based reviewers, however, face two structural limitations. First, they map manuscripts direct
arXiv:2608.00004v1 Announce Type: new Abstract: Grading natural-language mathematical proofs is a recurring cost in evaluating math-reasoning systems, and frontier LLM judges are expensive. We ask whether cheap open-weight models can serve as reliable judges given a candidate proof, a ground-truth
“世界模型，重点不是渲染的画质是否好看，而是模拟器是否能忠实还原现实物理规律和行为的因果”。11月3日，在深圳举行的百度AI Day·Apollo专场上，百度Apollo首次披露面向物理AI的智能体驱动全模态世界模型Apollo World，并提出世界模型 “2+3”判断标准。目前，该模型已用于萝卜快跑全球化测试与验证，助力无人驾驶规...
点击查看原文>
模型能够成功接入，并不代表调用体验已经足够顺畅。 在实际使用中，一个 API 模型往往不仅由算法工程师维护，还会被系统开发、业务分析等不同角色调用。输入参数较多时，需要反复对照接口文档填写；返回结果包含多层 JSON 结构时，又需要额外确认每个字段具体代表什么。 这些问题不会直接导致模型“不能用”，但会增加模...
手工压花器、二手 TITAN V，以及派友们用开源项目与 AI Coding 替代付费订阅的讨论。查看全文
点击查看原文>
今年双十一，购物车里最特别的一件商品，是一种看不见、摸不着的东西——Token（词元）。天猫 9 月 3 日上线 Token 充值中心，接入 DeepSeek、阿里云、智谱、Kimi，开售首日 Tok
GFlow（极风工作流平台）基于 RuleGo 规则引擎构建，核心理念是：AI 先审 · 人再签 · 签完自动办——大模型先审风险，低风险自动直通；会签、加签、任意退回，中国式审批一个不少；审批通过那一刻，入账、通知、回写业务系统由规则链自动执行。 本次 v1.4.1 围绕流转留痕、IM 集成、开放集成、表单打印、安全合规五大方...
点击查看原文>
普林斯顿大学计算机科学教授 Elad Hazan 提出利用“逆向博弈论”（机制设计）来约束高能力 AI 模型，通过借鉴人类社会的法律、监管与奖惩机制，让遵守规则成为 Agent 最理性的选择。该方案设计了“求解者”与“审计者”两类相互博弈与监督的 AI 角色，并辅以激励和复核机制防止串通作弊。然而，这一构想依赖于“AI 会做出理性行为”的核心假设，而当前大语言模型展现出的复杂与不可捉摸性使该假设面临挑战。这一探索为开发者和安全研究人员提供了基于制度经济学的新对齐思路，同时也揭示了控制复杂智能体所面临的机制边界。