समस्या
The softmax function turns a list of scores (logits) into probabilities. Language models use it to choose the next word:
softmax(x)_i = e^(x_i) / (e^(x_1) + e^(x_2) + … + e^(x_n))
Large scores make e^x overflow, so first subtract the largest score from every score. The result is mathematically the same.
Implement softmax(logits).
Input
logits: A list of scores.
Output
Return the list of probabilities, in the same order. They are all between 0 and 1 and add up to 1.
सीमाएँ
1 ≤ length ≤ 1,000
-1,000 ≤ each score ≤ 1,000
उदाहरण
इनपुट:
logits = [1.0, 2.0, 3.0]
अपेक्षित आउटपुट:
expected = [0.090031, 0.244728, 0.665241]
इनपुट:
std::vector<double> logits = {1.0, 2.0, 3.0};
अपेक्षित आउटपुट:
std::vector<double> expected = {0.090031, 0.244728, 0.665241};
इनपुट:
double[] logits = new double[] {1.0, 2.0, 3.0};
अपेक्षित आउटपुट:
double[] expected = new double[] {0.090031, 0.244728, 0.665241};
व्याख्या
e¹ ≈ 2.718, e² ≈ 7.389 and e³ ≈ 20.086 add up to 30.193, so the probabilities are about [0.090031, 0.244728, 0.665241].
इनपुट:
logits = [1000.0, 1000.0]
अपेक्षित आउटपुट:
expected = [0.5, 0.5]
इनपुट:
std::vector<double> logits = {1000.0, 1000.0};
अपेक्षित आउटपुट:
std::vector<double> expected = {0.5, 0.5};
इनपुट:
double[] logits = new double[] {1000.0, 1000.0};
अपेक्षित आउटपुट:
double[] expected = new double[] {0.5, 0.5};
व्याख्या
Computing e^1000 directly overflows. Subtracting the maximum gives [0, 0], and e⁰ / (e⁰ + e⁰) = 0.5 for both.
कैसे हल करें
पहले खुद कोशिश करें, फिर एक-एक करके संकेत खोलें।
संकेत 1
Subtract the largest logit before exponentiating so the largest transformed score becomes zero and overflow is avoided.
संकेत 2
Compute each exponent, sum them once, then divide every exponent by that total to get probabilities.
संकेत 3
Make sure the returned values stay in the same order as the input and still sum to one after normalization.
अपेक्षित जटिलता
O(n) time, O(n) extra space
संदर्भ
यह चुनौती CodeUtility के लिए अनुकूलित की गई है। ऊपर दिए गए स्टेटमेंट और सीमाओं का पालन करें।