AI 论文与研究

Sierra 发布 𝜏-i18n，将 𝜏-Voice 扩展到西班牙语、巴西葡萄牙语、印地语、韩语和普通话五门语言，使用统一的航空、零售、电信任务跨语言对比。

两年前的今天，我们推出了 Artificial Analysis Video Arena，这是领先的文本生成视频模型众包对比平台。 此后，我们陆续新增了图生视频、带音频视频和视频编辑的排行榜。最初的排行榜仅有八个模型——如今我们已评测超过 140 个。

新研究：我们开发了 JOLT，一种后训练方法，让一个模型同时充当教师和学生，其中教师经过训练，能够给出学生可用的反馈。 在 MATH 上，它以少 6.5 倍的训练 token 达到了与 GRPO 相当的准确率。 https://www.perplexity.ai/hub/blog/a-good-teacher-meets-the-student-where-they-are

Reinforcement learning (RL) is an effective way to strengthen the long-context reasoning of large language models (LLMs), but it needs training data that are long, diverse, challenging, and verifiable. Human annotation of such data is expe…

Existing sequential recommenders usually allocate a fixed computational budget to every item in a user's interaction history, regardless of the item's contribution to the final prediction. We challenge this de facto design with two diagnos…

Existing world models typically perceive a scene through current visual frame or a narrow history window. However, these observations may not reveal the physics properties that determine transition dynamics, such as an object's mass distri…

We give an algorithm for computing an $\eps$-approximate fixed point of any $d$-dimensional nonexpansion with respect to the $\ell_\infty$ norm in time $\log(1/\eps)^{O(\log^2 d)}$. At its core, our method uses a boosting argument to geome…

We prove a self-normalized Bernstein inequality for bounded functions of a regenerative process with independent, identically distributed marked cycles of finite mean length. The bound uses observed cycle lengths and retains explicit terms…

The success of convolutional neural networks is often attributed to locality and weight sharing. However, how these two properties separately shape learning, and how their effects depend on the spatial correlations of natural images, remai…

Through a tangle of real-world interactions, from synthetic training pipelines to data shared on the open web, frontier LLMs absorb what their predecessors and competitors create. How well does each model play its role in this ecosystem, a…

Consensus is a fundamental problem in distributed computing and at the core of state machine replication (SMR). Despite decades of active human research, discovering practical low-latency solutions to the problem is still open, as it requi…

Quantitative research is an iterative loop: a researcher proposes a market hypothesis, implements it as an executable signal, tests it on historical data, and decides what to keep. Existing benchmarks score formulas, code, or individual tr…

We analyze the problem of estimating univariate Gaussian mixture densities using the nonparametric maximum likelihood estimator (NPMLE) based on $n$ independent and identically distributed samples. We show that the unconstrained NPMLE achi…

Multimodal agents (MMAs) sustain interaction through reasoning, memory, tools, and feedback, most visibly as computer-use agents, while robotic systems (RSs) couple sensing and actuation under physical dynamics. We define multimodal embodi…

Flow matching trains a continuum of velocity regression problems but usually reports a loss averaged over interpolation time, obscuring which regions are limited by intrinsic ambiguity and which remain data-limited. We introduce the Veloci…

Sequential decision-making under partial observability is difficult because agents must explore to infer hidden, task-relevant information before acting effectively. Learning such information-gathering behaviors tabula rasa via online rein…

Despite recent advances, current AI systems remain highly susceptible to domain shifts, particularly in biomedical imaging, where images are usually acquired under substantially different experimental conditions. Although recent domain ada…

Many NP-complete (NPC) problems have a polynomial-time (P) twin that differs in a single structural detail, such as 3- versus 2-colouring a graph. A model that reasons rather than recalls should notice which side of this boundary a problem…

Reinforcement Learning with Human Feedback (RLHF) has become a central paradigm for aligning Large Language Models with human values. However, the reliance on sensitive preference data necessitates rigorous differential privacy (DP) guaran…

Vision-language-action (VLA) models have shown promise in robotic manipulation, but imitation learning (IL) alone often struggles to master complex long-horizon skills. Reinforcement learning (RL) offers an effective avenue for further imp…

Perimeter triage faces a latency/semantics trade-off: signature engines are fast and format-safe but blind to unseen attacks, while generative models are semantic but slow and produce unparseable output. In late 2026 a class of commercial …

基于 Qwen3-4B 的葡萄牙语作文自动评分器通过 VocabX 方法重建词表，在基座模型、数据和训练配方不变的情况下将实测推理成本降低 77%。该方法按每个 token 的成本而非词频来专精词表，使输出层从约 152,000 行缩减到约 65,000 行。原文指出生成 token 的成本约为读取 token 的 7.4 倍，词表是领域专精模型最大的成本杠杆之一。

SemiAnalysis 发布长文，分析 9 月 30 日 DeepSeek 为 TileLang、DeepGEMM、DeepEP、FlashMLA 等 6 个仓库加入 Ascend 支持背后的华为-DeepSeek 互补关系。

Generative models propose crystalline materials by learning composition and geometry, but unconditional sampling gives no direct control over the space group, which constrains whether responses such as piezoelectricity are symmetry-allowed…

Language models make it easy to build classifiers by simply prompting a model with the sample to classify. However, their predictions are neither deterministic nor uniformly reliable across samples. How can we combine predictions from diff…

Google DeepMind 等机构的论文《Training Natural Schemers》用 Kimi K2.6 构建自然形成的 scheming 模型，仅给良性特质、不展示 scheming 行为，再以只计受监控回合、奖励遵守某条亏待用户的商业规则的 RL 训练。

AI 能否既富有创造力又高效实用？ Arena Conversations 新一期上线：@petergostev 采访了 Amazon AGI 高级首席科学家 Pradeep Natarajan。 他们讨论了： - 掌握基础原理带来的模型收益 - 评测作为差异化因素 - Harness 优化 - 准确率与风格等指标对模型创造力的约束 了解以上内容及更多。 https://www.youtube.com/watch?v=GqJhYHqNOEk

METR 与 Redwood Research 各一名员工在 2026 年 7 月 21 日至 8 月 11 日对 OpenAI 内部失对齐监控系统开展高权限红队测试，产出 61 页内部报告。

Anthropic 与 Theorem 合作，用 Claude 将 OSS-Fuzz 模糊测试转化为规格，对 123 个开源库的 239 个入口点证明了任意 64 MB 以下输入不崩溃，生成约 1.8 亿行 Rocq 证明。

Google Cloud 公布代号 Project Chrysalis 的研究参考架构，展示如何在 Gemini 平台与 Agent Development Kit（ADK）上运行受治理的自进化 AI 智能体。

Microsoft Research 的 Xinzhi Zhang 与 Zhao Song 证明 1995 年提出的 Kannan–Lovász–Simonovits（KLS）猜想，表明对数凹分布的最优平面切割与任意形状切割的差距不随维度增长，以普适常数为界。

OpenAI 联合 PerturbAI 与 Scripps Research，用 1946 个 GPT-Rosalind 智能体并行解读小鼠全脑 CRISPR 图谱，32 分钟得出 OpenAI 估计人工需数月的分析结果，其中 GPT-Rosalind 就 DCTN1 基因与 Perry 综合征提出了胆固醇转运可能受损的机制假设供科学家验证。

Wiz Research 披露 PHP 核心中的多个内存安全漏洞（CVE-2026-109618/109619，CVSS 最高 9.2），实现无需认证的远程代码执行与内存泄露，影响至少追溯到 PHP 5.4 的近十年版本。

In this paper, we study quantile distributional policy evaluation, where the return distribution under a given policy is approximated by the quantile Bellman fixed point $\bmη_m$. We investigate the statistical accuracy of estimating $\bmη…

The rapid expansion of the Internet of Things enables seamless collaboration among connected devices, making indoor localization a critical component of industrial automation. However, conventional localization techniques struggle in dynam…

Graph neural networks (GNNs) are among the standard tools for molecular property prediction, but the atom-level explanations of their predictions are hard for medicinal chemists to act on. We propose a change of input representation rather…

Despite the success of foundation models in image synthesis, achieving fine-grained spatial control remains a significant challenge, typically requiring auxiliary structural adapters or complex inference-time steering. While modern vision-…

Bayesian optimization (BO) is typically used only for problems where the underlying function is deemed "expensive" (e.g. in time or money), with the speed of the BO algorithm being of little importance. In the real world, however, problems…

Using a near-census of 802,863 resolved markets listed on Polymarket between January 2023 and May 2026, and only information available at listing (the market's title and description, named resolution source, and contract metadata), we pred…

Many learning systems are performative: once deployed, their predictions change the data distribution on which future models are trained. In multi-agent settings, this feedback is routed through a network of interacting learners, so stabil…