WorldSonus: Bringing Sound to Worlds
Recent advances in world models have enabled increasingly realistic visual synthesis. However, these generated environme…
arXiv 人工智能预印本（Atom）
2026-10-07 01:50
未评分
反馈记录（521 条）
只显示我已产生过反馈的信息VeriFine: Scaling Verification for Self-Improvement in Embodied Reasoning
Self-improving policies continually expose new failure patterns, changing what their judges must be able to verify. Howe…
arXiv 人工智能预印本（Atom）
2026-10-07 01:50
未评分
AdvSim2Real : Training Web Agents Against Adaptive Prompt Injection in a Web World Model
Web agents complete user requests by reading and acting on pages that third parties write, so an instruction planted on…
arXiv 人工智能预印本（Atom）
2026-10-07 01:56
未评分
Agent in a Bottle: Can LLM Agents Turn Their Capabilities Into Cheap, Scalable Artifacts?
Large language models (LLMs) can solve many narrow tasks, but querying them separately for millions of related instances…
arXiv 人工智能预印本（Atom）
2026-10-07 01:57
未评分
Sherpa: Teaching LLMs to Teach Adaptively
Large language models (LLMs) have become increasingly capable problem solvers, but being able to solve a problem is not…
arXiv 人工智能预印本（Atom）
2026-10-07 01:58
未评分
DepthWorld: 3D World Model for Robot Manipulation
World models offer a data-driven alternative to traditional simulators for robotics, with applications spanning policy e…
arXiv 人工智能预印本（Atom）
2026-10-07 01:59
未评分
IdeaAnchor: Teaching LLMs to Turn Literature into Research Ideas
Scientific research often begins by synthesizing ideas from a set of related papers to identify gaps and formulate new d…
arXiv 人工智能预印本（Atom）
2026-10-07 01:59
未评分
4D-HOF: Hand-Object Flow Matching for Feed-Forward 4D Interaction Reconstruction
Existing methods for 4D hand-object reconstruction often rely on costly per-sequence optimization, while generative appr…
arXiv 人工智能预印本（Atom）
2026-10-07 01:59
未评分
What happens in the brain when cannabis makes you anxious?
Cannabinoids may intensify anxiety by removing a natural brake on specific neurons in the brain’s fear center, particula…
ScienceDaily 科研新闻（RSS）
2026-10-07 09:26
未评分
Animals may have evolved 200 million years earlier than we thought
New research suggests animals could have evolved as early as 800 to 700 million years ago, roughly 200 million years ear…
ScienceDaily 科研新闻（RSS）
2026-10-07 10:29
未评分
A new way to fight gum disease without wiping out good bacteria
Scientists discovered that disrupting the chemical messages bacteria use to communicate can push dental plaque toward a…
ScienceDaily 科研新闻（RSS）
2026-10-07 11:55
未评分
Humans may not be as hardwired for violence as we thought
A study comparing 100 primate species, including humans, suggests that everyday aggression and lethal violence did not e…
ScienceDaily 科研新闻（RSS）
2026-10-07 11:22
未评分
Do I Need the Cloud? Uncertainty-Aware Step-Level Handoff for Small Language Model Agents
Small language models (SLMs) are attractive as local agent controllers because they reduce remote inference, latency, an…
为什么值得关注：属于技术进展，可能扩散到其他领域
arXiv 人工智能预印本（Atom）
2026-10-06 14:09
综合价值 3.8 / 5
One Step at a Time: Trading LLM Autonomy for Process Predictability
Organizations automating operational processes need more than a correct outcome: they need to predict how a process will…
为什么值得关注：属于技术进展，可能扩散到其他领域
arXiv 人工智能预印本（Atom）
2026-10-06 14:10
综合价值 3.8 / 5
Agentic Semantic Sensing for Resource-Adaptive AI-RAN
Semantic sensing (SemS) acquires task-relevant information rather than reconstructing complete physical information.Exis…
为什么值得关注：提供了新的研究结果
arXiv 人工智能预印本（Atom）
2026-10-06 14:27
综合价值 3.9 / 5
Harness Engineering for Software Engineering via Modular Executable Dev-Primitives
Large language models (LLMs) equipped with terminal access have demonstrated strong capabilities in automating software…
为什么值得关注：属于技术进展，可能扩散到其他领域
arXiv 人工智能预印本（Atom）
2026-10-06 14:32
综合价值 3.8 / 5
DHCG: Dynamic Construction of Hierarchical Collaboration Graphs for LLM-Based Multi-Agent Reasoning
LLM-based multi-agent systems (MAS) have demonstrated strong capabilities in solving complex problems across diverse dom…
为什么值得关注：属于技术进展，可能扩散到其他领域
arXiv 人工智能预印本（Atom）
2026-10-06 14:33
综合价值 3.8 / 5
Dynamic Positional Attention Modulation for Parameter-Efficient Fine-Tuning of Large Language Models
Parameter-efficient fine-tuning (PEFT) has become a standard approach for adapting large language models to downstream t…
为什么值得关注：属于技术进展，可能扩散到其他领域
arXiv 人工智能预印本（Atom）
2026-10-06 14:55
综合价值 3.8 / 5
RA-MoWE: Workflow-Affinity Embeddings for Query Clustering and Agentic Workflow Generation
Agentic workflows enable large language models (LLMs) to solve complex tasks by coordinating reasoning, tool use, and ve…
为什么值得关注：属于技术进展，可能扩散到其他领域
arXiv 人工智能预印本（Atom）
2026-10-06 14:58
综合价值 3.9 / 5
WorkflowOps: Learning Agent Collaboration Priors for Multi-Agent Workflow Orchestration
Multi-agent systems are increasingly deployed for complex knowledge work, yet their orchestration layers remain largely…
为什么值得关注：属于技术进展，可能扩散到其他领域
arXiv 人工智能预印本（Atom）
2026-10-06 15:01
综合价值 3.9 / 5
近 30 天反馈概览
用于理解自己的注意力分配是否与预期一致
标记重要
4
标记一般
17
标记不重要
8
尚未评价
468
收藏总数
2
跟踪总数
3
按主题的价值判断分布
近 30 天，我标记为「重要」的比例
心理学
重要 2
一般 3
不重要 0
共 5
40%
教育改革
重要 1
一般 2
不重要 0
共 3
33%
教育
重要 2
一般 5
不重要 1
共 8
25%
脑科学与神经科学
重要 1
一般 4
不重要 1
共 6
17%
AI
重要 1
一般 4
不重要 5
共 10
10%
生物
重要 0
一般 3
不重要 1
共 4
0%
历史
重要 0
一般 1
不重要 2
共 3
0%
哲学
重要 0
一般 1
不重要 0
共 1
0%
主题权重建议
系统只做提示，最终决定权在你- 近 60 天你把「心理学」相关信息标记为「重要」的比例为 33%（样本 3 条），是否将主题权重从 1.5 调整为 0.6？（系统不会自动修改，需要你确认）
- 近 60 天你把「教育改革」相关信息标记为「重要」的比例为 33%（样本 3 条），是否将主题权重从 1.5 调整为 0.6？（系统不会自动修改，需要你确认）
- 近 60 天你把「教育」相关信息标记为「重要」的比例为 50%（样本 4 条），是否将主题权重从 1.5 调整为 0.9？（系统不会自动修改，需要你确认）
- 近 60 天你把「脑科学与神经科学」相关信息标记为「重要」的比例为 67%（样本 3 条），是否将主题权重从 1 调整为 1.3？（系统不会自动修改，需要你确认）