1
中国科学院青岛能源所开发出新型高容量与强选择性海水提铀材料
IT之家 9 月 2 日消息，中国科学院青岛生物能源与过程研究所（简称“青岛能源所”）宣布， 该所研究团队近期在海水提铀材料领域取得系列进展 。 团队首次将离散分子拓扑设计引入多孔有机笼体系，合成出磷酸功能化材料 PhosCage，并进一步将其与芳纶纳米纤维复合，制备出兼具高选择性、高容量和抗生物污…
2
I-CALM: Incentivizing Confidence-Aware Abstention for LLM Selective Answering
让大模型学会“不知道就承认”，用激励机制打造可靠的选择性回答，直击幻觉痛点。
arXiv:2604.03904v2 Announce Type: replace-cross Abstract: Large language models (LLMs) often produce confident but incorrect answers, in part because …
3
Asymptotic Risk Calibration for Selective Question Answering
不看全文也能“有把握放弃回答”？这项研究为选择性问答系统提供了风险校准的理论保障。
arXiv:2608.12008v1 Announce Type: new Abstract: Large language models (LLMs) may generate fluent but incorrect answers, making uncertainty quantificat…
4
Forget Total Recall. Give Your AI Agent Selective Memory with OpenSearch.
别让长文本淹没重点，用OpenSearch为AI智能体打造选择性记忆，精准调用所需信息。
Imagine handing an LLM the full text of Anna Karenina and asking what Levin thinks about farming. The model has the information somewhere in its windo…
5
Rethinking LLM Verification: Evidence Structure, Uncertainty, and Selective Refinement
跳出传统校验框架，用证据结构和不确定性指导选择性修正，给LLM验证提供了新思路。
arXiv:2608.10725v1 Announce Type: new Abstract: Large language models (LLMs) often rely on shortcuts rather than systematic reasoning, raising safety …
6
LLM-INSTRUCT at UZH Shared Task 2026: Constraint-Aware Retrieval and Selective Debate for Paragraph-Level Argument Mining
UZH共享任务2026论文：基于约束感知检索与选择性辩论的段落级论证挖掘新方法
arXiv:2607.20430v1 Announce Type: cross Abstract: We present LLM-INSTRUCT, the winning system for the UZH Shared Task at ArgMining 2026 on paragraph-l…
7
Calibrated Selective Fact-Checking via Evidence Chain Evaluation
LLM事实核查的二元判断隐藏可靠性危机，论文提出基于证据链评估的校准选择性方法，直击AI可信度痛点。
arXiv:2607.18240v1 Announce Type: new Abstract: Large language models (LLMs) can achieve strong fact-checking accuracy, yet forced binary decisions co…
8
Shared Selective Persistent Memory for Agentic LLM Systems
这篇arXiv论文提出面向Agentic LLM系统的共享选择性持久记忆机制，解决多智能体协作中的记忆共享与选择性持久化问题，是AI大模型研究前沿。
arXiv:2607.09493v1 Announce Type: new Abstract: Agentic LLM systems that generate code through multi-turn tool use face a fundamental context problem:…
9
Super Weights in LLMs and the Failure of Selective Training
揭秘LLM中隐藏的“超级权重”：为何选择性训练策略翻车？COLM 2026接受研究揭示深层原理。
arXiv:2607.08733v1 Announce Type: new Abstract: Recent work identified Super Weights, individual parameters whose removal degrades model performance b…
10
$C$-$\Delta\Theta$: Circuit-Restricted Weight Arithmetic for Selective Refusal
从电路层面微调权重，让大模型学会选择性拒绝——AI安全的新思路。
arXiv:2602.04521v2 Announce Type: replace Abstract: Modern deployments require LLMs to enforce safety policies at scale, yet many controls rely on inf…
11
Aligning Language Models with Selective Prediction
大模型在不确定时学会说“不知道”，选择性预测校准方法，提升安全性与可靠性。
arXiv:2607.03528v1 Announce Type: cross Abstract: Large language models (LLMs) are increasingly deployed as critical decision-making components in hig…
12
4nm近售罄、部分8nm接近满载 三星晶圆代工开始选择性接单
据报道，业内人士称，随着AI半导体需求升温、全球大型科技公司订单增加，三星电子晶圆代工部门已针对部分制程启动配额机制（allocation），将有限产能集中分配给较重要的客户和制程，将产能优先分配给现有客户，新客户订单则选择性承接。（新浪财经）
13
三星4nm产能基本售罄，部分8nm产能满负荷运转，已开始选择性接单
据ChosunBiz，业内人士透露，三星电子晶圆代工部门近期调整了供货策略，优先处理现有客户的订单，并有选择地接受新客户的订单。行业分析师指出，人工智能市场的爆炸式增长正在从根本上改变晶圆代工的需求结构。此前主要集中在智能手机应用处理器（AP）上的先进制造工艺需求，近期已转向人工智能加速器芯片、专用…
14
现在是供方时间：消息称三星晶圆代工已开始选择性接受新客户订单
IT之家 7 月 2 日消息，据韩媒《ChosunBiz》当地时间今日报道称，此前因产能利用率低下深陷亏损泥潭的 三星晶圆代工当前已在有选择性地接受新客户订单 ，显示晶圆代工市场已进入供方主导时间。 有消息指出，三星晶圆代工的主要先进制程工艺节点 4nm 今年的产能已基本售罄，甚至明年产能也已排满；…
15
Selective Memory Retention for Long-Horizon LLM Agents
ICML 2026收录，揭秘LLM智能体如何在长期任务中精选记忆、提升性能。
arXiv:2606.29178v1 Announce Type: cross Abstract: When does retention matter for memory-augmented LLM agents? We study this with TraceRetain, a lightw…
16
Check Yourself Before You Wreck Yourself: Selectively Quitting Improves LLM Agent Safety
让AI智能体学会“适时退出”，用选择性放弃大幅提升多轮工具调用场景的安全性。
arXiv:2510.16492v4 Announce Type: replace Abstract: As Large Language Model (LLM) agents increasingly operate in complex environments with real-world …
17
Safety Is Not Universal: The Selective Safety Trap in LLM Alignment
揭秘LLM安全评估的致命缺陷：看似全面防护实则暗藏针对特定群体的盲区，戳破“选择性安全陷阱”系统性风险
arXiv:2601.04389v3 Announce Type: replace-cross Abstract: Current safety evaluations of large language models (LLMs) create a dangerous illusion of un…
18
RepSelect: Robust LLM Unlearning via Representation Selectivity
大模型遗忘难题新解法：基于表征选择性实现鲁棒的知识去除，平衡保留与遗忘的边界。
arXiv:2606.17168v1 Announce Type: new Abstract: Making large language models (LLMs) deeply forget specific knowledge and values without sacrificing ge…
19
Mechanism-Guided Selective Unlearning for RLVR-Induced Reasoning
提出机制引导的选择性遗忘方法，精准消除RLVR引发的错误推理，保持模型原有能力。
arXiv:2606.19222v1 Announce Type: new Abstract: We propose MAST (Mechanism-Aligned Selective Targeting), a mechanism-guided method for unlearning RLVR…
20
Think-at-Hard: Selective Latent Iterations to Improve Reasoning Language Models
被ICML'26录用的新方法，通过选择性潜在迭代显著提升推理语言模型的表现力。
arXiv:2511.08577v3 Announce Type: replace-cross Abstract: Improving the reasoning abilities of Large Language Models (LLMs), especially under paramete…