491–500 of 500 posts
paywalled, but I've seen this in action.
Meaning you were pressured by an employer? Would be interesting if Glassdoor periodically reached out to users to revise reviews and capture sentiment "before and after" working at a given company.
Not necessarily under handed, but still gaming the system.
Earlier quoted context omitted.
It's remarkable how often companies expect their employees to be completely loyal and do a bunch of work for free, while also treating their employees like free labor and cutting them loose at the smallest disagreement. I've started steering very clear of any company that talks about how important loyalty is, or how the employees are all 'like a family'. Biggest red flags for toxic management imo.
> It's remarkable how often companies expect their employees to be completely loyal and do a bunch of work for free, while also treating their employees like free labor and cutting them loose at the smallest disagreement. There's lots of other imbalances as well like how you have to tolerate being mistreated so you can get a good reference, and you have to actively avoid mentioning negative reasons you left a company…
I've just been giving out former coworkers' contact info instead, and usually they don't even get called. I'm guessing they call the former employer just to verify I actually worked there, but that's it
By and large, GlassDoor ratings are no longer a good measure of how well a company treats employees; they now measure, mainly, whether a company has the ability to engineer and maintain artificially good GlassDoor ratings. GlassDoor, in short, has become a textbook example of Goodhart's Law: "When a measure becomes a target, it ceases to be a good measure."[a] The same phenomenon is known in some contexts as Campbell…
https://news.ycombinator.com/item?id=18364148
where you discuss with one of the authors of the excellent paper "Towards Understanding Linear Analogies"
you are discussing ELMo and specifically word senses i.e. "leaves" which has multiple senses (departs, foliage, ...)
I recently stumbled on a paper from 2016 (modified 2018) which IMHO gives a lot of insight, but I had to read both the old and the new version (I recommend reading v1 first and then the newest)
They illustrate how for example the word "leave" in word embeddings, is in fact simply a linear combination (with coefficients on the order of 1) of the true positions of each individual sense i.e. "leave" = A"leave.1"+B"leave.2"+... with A,B, ... constants close to 1. Theres typically less than 10 for a single word.
These reside in the same vector space as the word embeddings, and they illustrate how these sense vectors can be retrieved from the shallow word embedding vectors by sparse coding!
The paper is at https://arxiv.org/abs/1601.03764
Again I recommend reading first v1 then v6
Earlier quoted context omitted.
"may" is such a weasel word. If you're a cellular company, "may" means "will". "If you are in the higher tiers of data use, we may throttle your bandwidth". If you're Glassdoor who sells reputation management and HR services to the very employers being reviewed, "may" means "we're paying lip service to our critics".
May throttle doesn't necessarily mean "will throttle." The company is giving itself the latitude not to throttle, or to not throttle exactly at the data cap. In other words, you can't rely on that throttling if you're, say, connecting to some cloud service that's charging you by the gigabyte to transfer data.
Earlier quoted context omitted.
The negative review was informative enough. Political, negative work environment and old, dead tech. Blah blah. It's the context that is particularly damning. Whenever you see a heartfelt negative review surrounded by obviously fake or reactionary (do you really not see that?) positive reviews, that is a red flag. It is not uncommon.
I totally disagree. When I read reviews (on Glassdoor or anywhere, really) I try to discount any emotionally charged content and focus on the factual elements of the reviews. I mean, "Political, negative work environment"? Every single group of humans since the beginning of time has a level of political interaction, so when I see comments about things being political I pretty much discount them unless there are some…
By and large, GlassDoor ratings are no longer a good measure of how well a company treats employees; they now measure, mainly, whether a company has the ability to engineer and maintain artificially good GlassDoor ratings. GlassDoor, in short, has become a textbook example of Goodhart's Law: "When a measure becomes a target, it ceases to be a good measure."[a] The same phenomenon is known in some contexts as Campbell…
sorry, this is off topic, but it's impossible to add a comment to the discussion in which you participated at https://news.ycombinator.com/item?id=18364148 where you discuss with one of the authors of the excellent paper "Towards Understanding Linear Analogies" you are discussing ELMo and specifically word senses i.e. "leaves" which has multiple senses (departs, foliage, ...) I recently stumbled on a paper from 2016…
Earlier quoted context omitted.
sorry, this is off topic, but it's impossible to add a comment to the discussion in which you participated at https://news.ycombinator.com/item?id=18364148 where you discuss with one of the authors of the excellent paper "Towards Understanding Linear Analogies" you are discussing ELMo and specifically word senses i.e. "leaves" which has multiple senses (departs, foliage, ...) I recently stumbled on a paper from 2016…
Interesting. Based on a quick glance, it seems this would answer the question I asked in that thread about whether it might be possible to get word-sense embeddings via two simpler transformations: first a transformation to the space of word-sense compositions (e.g., via GloVe/SGNS), and then a transformation to the space of word senses. I'll take a closer look. Thank you!
corpus -- word2vecOrGloVe--> word embeddings v_w in R^n
word embedding --sparsecoding--> sense embeddings v_s in same R^n
the sparse coding process gives the constants A_ws and senses where subscript w is a word index and s is a sense index, so that:
word vectors v_w = sum(A_ws v_s, s)
and for each word w most A_ws are zero except for a few s values
1) polysemy: a word w can have multiple senses, namely those sense vectors with index s where A_ws is nonzero
2) synonyms: a sense s can have multiple synonyms w, again those w where A_ws is nonzero
so the result of sparse coding gives for each word, a couple of indexes of the sense vectors, and for each sense the corresponding indexes of word vectors... and of course the sense vectors themselves.
so that to find say a synonym of "leaves", you just look at the sense indexes corresponding to that string, then you look at the different words indexes for that sense, and they will refer to the words "foliage" but also "leaves" of course and possibly others...
I also believe that once you have the sense vectors, in theory a second pass through the corpus should improve results if the context of each focus word is used to determine the closest sense vector compatible with the focus word... so that in effect word2vec or Glove extraction is run on the senses instead of the words
for the sparse coding they used SMALLbox, and I am still trying to better grasp how exactly the sparse coding works, and what prevents the A_ws and v_s to reduce to the trivial solution v_s = v_w and A_ww = 1 and A_wx = 0 for x differing from w...
https://twitter.com/marsxrobertson/status/108837592864696729...
Good rankings help them hire, why wouldn't they optimise the ranking?
Anyone have a good alternative to Glass Door. Googling the company is probably the easiest but then you run the risk of data biased the other way.