Signal Brief

二年级学生Jo Nagai发现蝴蝶可遗传记忆

日本东京都二年级学生Jo Nagai注意到养育的食蚜 caterpillars 在蝴蝶化后仍保持对薰衣草的回避行为,经Georgetown大学Entomologist Dr. Martha Weiss合作完成实验:70%训练过的蝴蝶及其后代均表现出对薰衣草的遗传性回避,记忆在全变态发育中保存并遗传。

twitter关注列表 ᴅᴀɴɪᴇʟ ᴍɪᴇssʟᴇʀ 🛡️ (@DanielMiessler) 发布 2026-07-18 收录 2026-07-18 值得跟进

一句话判断

首次实验证明哺乳动物记忆在全变态生物中的保存及遗传机制,挑战传统神经科学共识,具备突破性理论价值

核心信息

日本东京都二年级学生Jo Nagai注意到养育的食蚜 caterpillars 在蝴蝶化后仍保持对薰衣草的回避行为,经Georgetown大学Entomologist Dr. Martha Weiss合作完成实验:70%训练过的蝴蝶及其后代均表现出对薰衣草的遗传性回避,记忆在全变态发育中保存并遗传。

原始内容

ᴅᴀɴɪᴇʟ ᴍɪᴇssʟᴇʀ 🛡️ (@DanielMiessler) 转发了 Tim Ferriss (@tferriss) 的帖子: This makes me happy on so many levels. > **引用原帖 Carolyn Rockey (@CarolynWRockey):** > Jo Nagai was raising swallowtail butterflies at his home in Kobe, Japan, when he noticed something odd. The ones he had looked after as caterpillars seemed to recognize him. Wild butterflies fled. His didn't. > He was in second grade. He wrote a four-page letter to Dr. Martha Weiss, an entomologist at Georgetown University who had studied whether moths could retain memories through metamorphosis. He asked if she could help him design a version of her experiment for butterflies. > She said yes. > Using a muscle therapy device, Jo trained caterpillars to associate the scent of lavender with a mild vibration. When the caterpillars became butterflies, 70 per cent of them still avoided the lavender. Their brains had been completely rebuilt during metamorphosis. The memory survived anyway. > Then he bred them. > The offspring, which had never been trained, also avoided lavender. So did their grandchildren. Without ever experiencing the vibration, two generations of butterflies inherited an aversion to a scent their grandmother had been taught to fear. > Jo documented it all in a 33-page research paper and presented his findings at the International Congress of Entomology in Kobe in 2024. He was 10. > A second grader wrote a letter to a Georgetown professor, and together they found evidence that butterflies can pass memories down through generations. > -Wilderness Whisper > https://x.com/CarolynWRockey/status/2077742993658851529

相关动态

01

Kimi K3 在 SpreadsheetBench 2 上排名第一

Kimi K3 在 SpreadsheetBench 2 基准测试中排名第一,超越 Claude Fable 5,完成 34.8% 的 workflow 任务。该基准测试涉及 321 个专家策划任务,平均每个任务包含 11.8 个工作表和 593.5 个单元格更改,聚焦于完整的电子表格工作簿执行。

twitter关注列表2026-07-18#模型#技术突破#评测
观察
02

Mind Lab 开源长文本强化学习项目

Mind Lab 开源了一个长文本强化学习(RL)项目,支持 2M tokens 的长上下文。相比以往需要数千个 GPU 的 1M 上下文研究,该项目仅需 8 个 GPU 即可完成,大幅降低了超长上下文研究的算力门槛。

twitter关注列表2026-07-17#开源#技术突破#模型
观察
03

Kimi K3非对称竞争分析

Kimi K3在Arena前端/WebDev人类盲测中以1679 Elo排名第一,领先Fable 5(1631)和GPT-5.6 Sol(1618),但在综合智能指数中仅排第四(57.1分),落后Fable 5(59.9)和GPT-5.6 Sol(58.9)。K3定价15美元/百万输出tokens,远低于Fable 5的50美元,并计划7...

twitter关注列表2026-07-18#AI#模型#技术突破
观察
04

AI 代码生成大势所趋

Greg Isenberg 发文指出,相比一年前的手写代码实践,如今大多数工程代码已由 AI 生成,标志着编程范式的根本转变。他援引了 Google 75% 新代码由 AI 生成、Anthropic 90%+ 代码由 Claude 编写、GitClear 代码重复率上升 81% 复用率下降 70% 等具体数据,并引用 Dario Amod...

twitter关注列表2026-07-18#技术突破#行业动态#分析
观察
05

开源模型长期网络能力差距缩小至4-7个月

长期网络能力方面,领先开源模型落后闭源前沿从2025年大部分时间的6-10个月缩短至4-7个月。GLM-5.2匹配了约4个月前发布的闭源模型。AI安全研究所的32步“The Last Ones”任务中,GPT-5.6 Sol在10次尝试中完成7次,每次预算1亿token,且性能随推理token增加而提升。

twitter关注列表2026-07-17#模型#技术突破#AI安全
观察