Interview Questions

Exponent - "This is another Diagnosis problem. To answer this question, we suggest you use our framework (along with the TROPIC method) to be as thorough as possible. The framework is as follows: Ask clarifying questions List potential high level reasons Gather Context (TROPIC)Time Region Other features / products (internal) Platform Industry / Competition Cannibalization Establish a theory of probable cause Test theories Propose solutions Summarize "See full answer

Anonymous Hornet - "label #shape = [batchsize, posids] X #shape = [batch_size, ch] hidden = nn.Linear(batchsize, ch) #shape = [batchsize, ch] logits = nn.Linear(ch, C) #shape = [ch, C] #modifications start here neg_ids = torch.randint(0, C, 0.1 * C) #e.g. if we want to randomly sample only 10% of C outputneg = np.dot(logits[:, negids], hidden(X)) #shape [batch_size, C1] outputpos = np.dot(logits[:, posids], hidden(X)) #shape [batch_size, C2] output = torch.concat(outputpos, outputneg) #concatenate out"See full answer

Exponent - "This is an Improve a Product question. Let's first go over the Improve a Product formula: Ask clarifying questions Identify users, behaviors, and pain points State product goal Brainstorm small improvements Brainstorm bolder improvements Measure success Summarize Now, let's begin! Ask clarifying questions Before we begin listing off recommendations, it's important you ask questions to ensure you and the interviewer are on the same page"See full answer

Alex W. - "Memory allocation happens for storing a reference pointer as well as the default size of the generic object class depending on the language this is called in. Assuming this is in a JVM, this data is stored in metaspace, and memory allocation happens in heap."See full answer