TE
科技回声
首页24小时热榜最新最佳问答展示工作
GitHubTwitter
首页

科技回声

基于 Next.js 构建的科技新闻平台,提供全球科技新闻和讨论内容。

GitHubTwitter

首页

首页最新最佳问答展示工作

资源链接

HackerNews API原版 HackerNewsNext.js

© 2025 科技回声. 版权所有。

Enabling Continual Learning in Neural Networks

152 点作者 interconnector大约 8 年前

5 条评论

cs702大约 8 年前
This builds on the ideas behind PathNet, previously discussed at <a href="https:&#x2F;&#x2F;news.ycombinator.com&#x2F;item?id=13675891" rel="nofollow">https:&#x2F;&#x2F;news.ycombinator.com&#x2F;item?id=13675891</a><p>Whereas PathNet permanently freezes parameters and pathways used for previously learned tasks, in this case the authors compute how important each connection is to the most recently learned task, and protect each connection from future modification by an amount proportional to its importance. Important pathways tend to persist, and unimportant pathways tend to be discarded, gradually freeing &quot;underused&quot; connections for learning new tasks.<p>The authors call this process Elastic Weight Consolidation (EWC). Figure 1 in the paper does a great job of explaining how EWC finds solutions in the search space of solutions that are good for new tasks without incurring significant losses for previous tasks.<p>Very cool!
评论 #13874886 未加载
rayuela大约 8 年前
I&#x27;m confused. I don&#x27;t get what the novelty in this is. It looks like all they do is include an input that identifies different tasks and then trains one neural network to learn a separate distributions for each task, with some weight sharing...
评论 #13871110 未加载
评论 #13871142 未加载
colmvp大约 8 年前
Sidenote: On the list of contributors I noticed there are Research Engineers and Research Scientists. What is the difference between the two?
评论 #13872477 未加载
apl大约 8 年前
This came out two days ago and uses what they call intelligent synapses to improve multi-task learning: <a href="https:&#x2F;&#x2F;arxiv.org&#x2F;abs&#x2F;1703.04200" rel="nofollow">https:&#x2F;&#x2F;arxiv.org&#x2F;abs&#x2F;1703.04200</a><p>Seems closely related.
spynxic大约 8 年前
&quot;..After learning a task, we compute how important each connection is to that task.&quot;<p>Anyone know if this was expanded on in the whitepaper?
评论 #13874281 未加载