|
- intuition - What is perplexity? - Cross Validated
So perplexity represents the number of sides of a fair die that when rolled, produces a sequence with the same entropy as your given probability distribution Number of States OK, so now that we have an intuitive definition of perplexity, let's take a quick look at how it is affected by the number of states in a model
- 求通俗解释NLP里的perplexity是什么？ - 知乎
所以在给定输入的前面若干词汇即给定历史信息后，当然语言模型等可能性输出的结果个数越少越好，越少表示模型就越知道对给定的历史信息 \ {e_1\cdots e_ {i-1}\} ，应该给出什么样的输出 e_i ，即 perplexity 越小，表示语言模型越好。
- 如何评价perplexity ai，会是未来搜索的趋势吗？ - 知乎
Perplexity AI 不是搜索的终点，但可能是我们逃离“信息垃圾场”的起点。 它就像是搜索引擎界的 GPT-4：懂你说什么，还知道去哪儿找答案。
- 除了Newbing、WallesAI、Perplexity之外，还有哪些好用的AI搜索引擎? - 知乎
像Perplexity处理一个问题，开启 co-pilot之后会引导你问下一个问题，但是秘塔搜索是直接把问题的所有可能性都展示给你并且还搭配了思维导图了大纲。 另外，它最近更新的一款学术模式，在专业度来说比Perplexity这些是好很多的。
- Comparing Perplexities With Different Data Set Sizes
Would comparing perplexities be invalidated by the different data set sizes? No I copy below some text on perplexity I wrote with some students for a natural language processing course (assume log log is base 2): In order to assess the quality of a language model, one needs to define evaluation metrics One evaluation metric is the log-likelihood of a text, which is computed as follows
- clustering - Why does larger perplexity tend to produce clearer . . .
Why does larger perplexity tend to produce clearer clusters in t-SNE? By reading the original paper, I learned that the perplexity in t-SNE is 2 2 to the power of Shannon entropy of the conditional distribution induced by a data point
- 知乎 - 有问题，就会有答案
知乎，中文互联网高质量的问答社区和创作者聚集的原创内容平台，于 2011 年 1 月正式上线，以「让人们更好的分享知识、经验和见解，找到自己的解答」为品牌使命。知乎凭借认真、专业、友善的社区氛围、独特的产品机制以及结构化和易获得的优质内容，聚集了中文互联网科技、商业、影视
- machine learning - Why does lower perplexity indicate better . . .
The perplexity, used by convention in language modeling, is monotonically decreasing in the likelihood of the test data, and is algebraicly equivalent to the inverse of the geometric mean per-word likelihood A lower perplexity score indicates better generalization performance I e, a lower perplexity indicates that the data are more likely
|
|
|