微信内可能无法直接打开本站。请点右上角 ··· → 在浏览器打开，或复制链接后用系统浏览器访问。
XMT
信流 · 上滑连读 · 来源可核
Factcheck: 10 Flaws in the Conservative Report on “Cheap Power”
By Josh Gabbatiss, Molly Lempriere, Simon Evans In a new report, the opposition Conservatives argue that UK electricity prices are too high and that it would be better for the climate to have cheap electricity, even if that means using more gas.…
- The idea is that cheap power would encourage people to use .
- [continued] The post Factcheck: 10 Flaws in the Conservative Report on…
RSS 官方收录 · 可信分层展示
世界模型进入下半场：自变量 WALL-SS 突破三大瓶颈，让虚拟世界成为机器人的「训练场」
机器人世界模型的下一场竞争，可能不是画质，是架构。作者｜Li Yuan 编辑｜郑玄 一只机械臂的夹爪明明没有夹住杯子，杯子却像被磁铁吸住一样，跟着夹爪升了起来。这个看似荒诞的问题，正在困扰机器人世界模型，业内甚至将其称为「磁铁式抓取」。…
- 视频生成模型只要让画面看起来合理，就算完成任务；但世界模型不能只追求「像真的」。
- 机器人训练数据大多来自成功演示，模型容易形成捷径：与其理解动作的细微差别，不如按照画面，直接生成最可能的结果。
- 可在真实世界里，夹爪位置偏移几毫米，就可能让结果从抓起变成碰倒杯子甚至抓空。
RSS 官方收录 · 可信分层展示
功夫朋克，跃然眼前：VITURE ×《影之刃零》收藏级联名 XR 眼镜科隆全球首发，今日开启预购
2026 年 8 月 28 日，全球 XR 眼镜领导品牌 VITURE 携手《影之刃零》开发商灵游坊（S-GAME），于科隆游戏展期间举办的腾讯十周年全球艺术展上，正式揭晓收藏级联名之作——VITURE ×《影之刃零》Phantom Beast「幻影猛兽」XR 眼镜。…
- 这不仅是一款联名产品，更是一件可以佩戴的典藏作品：它将《影之刃零》独树一帜的「功夫朋克」美学融入 VITURE 旗舰级 XR 技术，让玩家透…
- 联名眼镜即日起于 VITURE 京东自营旗舰店独家开启预购，首发价 4,299 元，并将于 10 月 29 日与《影之刃零》同步发售、正式发货。
- 预售期间下单，还可获赠独家珍藏版游戏地图卷轴。
RSS 官方收录 · 可信分层展示
AI Revealed Preferences
arXiv:2608.26178v1 Announce Type: new Abstract: There is growing interest in whether language models have stable preferences, for technical, safety, and philosophical reasons.…
- We test 20 language models and find a range of preferences---stable di…
- We run three forced-choice experiments on revealed rather than stated …
- Headline findings include evidence that models are tedium-averse, "lei…
RSS 官方收录 · 可信分层展示
Knowledge Cards: Structured Knowledge for AI Systems
arXiv:2608.…
- 26176v1 Announce Type: new Abstract: AI systems whose outputs inform r…
- Established documentation artefacts already capture important aspects …
- Model cards describe how a system behaves, data cards describe what it…
RSS 官方收录 · 可信分层展示
Refusal Is Not Robustness: Auditing Confident Fabrication in Large Language Models on a Provably Uninformative Clinical Pain Speech Transcript
arXiv:2608.…
- 26167v1 Announce Type: new Abstract: Hallucination and abstention benc…
- Seven large language models were evaluated on the TAME Pain speech cor…
- Participants read phonetically balanced Harvard Sentences while one ha…
RSS 官方收录 · 可信分层展示
A Task-Centric Ontology and Deterministic Domain Rules as a Verifiable Core for AI-Assisted Chemistry Problem Solving
arXiv:2608.…
- 26164v1 Announce Type: new Abstract: Large language models can interpr…
- This paper presents ChemOntoRule, a proof-of-concept symbolic core for…
- The central design choice is task-centric ontology engineering: the on…
RSS 官方收录 · 可信分层展示
A Safety-Gated Multimodal AI Backend for Mental-Health Support: Hierarchical State Representation, Conservative Risk Fusion, and Controlled Generation in Anian
arXiv:2608.…
- 26162v1 Announce Type: new Abstract: Safety-critical mental-health sup…
- This paper presents Anian, a safety-gated multimodal AI backend for pe…
- Anian is not intended to diagnose psychiatric conditions or replace cl…
RSS 官方收录 · 可信分层展示