From 6311447e2640b52b4a6b3d5dac99b8b5dfd848ea Mon Sep 17 00:00:00 2001
From: Yuval Levental <yhl3051@rit.edu>
Date: Tue, 9 Dec 2025 13:53:12 +0000
Subject: [PATCH] (Type what you did)

---
 .../tensor-operations/terms/sign/sign.md      | 62 +++++++++++++++++++
 1 file changed, 62 insertions(+)
 create mode 100644 content/pytorch/concepts/tensor-operations/terms/sign/sign.md

diff --git a/content/pytorch/concepts/tensor-operations/terms/sign/sign.md b/content/pytorch/concepts/tensor-operations/terms/sign/sign.md
new file mode 100644
index 00000000000..17a209b5c61
--- /dev/null
+++ b/content/pytorch/concepts/tensor-operations/terms/sign/sign.md
@@ -0,0 +1,62 @@
+---
+Title: '.sign()'
+Description: 'Returns a tensor with the sign of each element, indicating whether it is negative, zero, or positive.'
+Subjects:
+  - 'Computer Science'
+  - 'Data Science'
+Tags:
+  - 'Deep Learning'
+  - 'Methods'
+  - 'Programming'
+  - 'PyTorch'
+CatalogContent:
+  - 'intro-to-py-torch-and-neural-networks'
+  - 'paths/data-science'
+---
+
+The **`.sign()`** method in PyTorch returns a new [tensor](https://www.codecademy.com/resources/docs/pytorch/tensors) with the sign of each element from the input tensor. It returns -1 for negative values, 0 for zero, and 1 for positive values. This method is commonly used in gradient-based optimization, activation functions, and mathematical operations where the direction or polarity of values matters.
+
+## Syntax
+
+```pseudo
+torch.sign(input, *, out=None)
+```
+
+**Parameters:**
+
+- `input` (Tensor): The input tensor whose signs are to be computed.
+- `out` (optional): A tensor to store the output. Must have the same shape as `input`.
+
+**Return value:**
+
+The `.sign()` method returns a new tensor containing the sign of each element in the `input` tensor. Unless the `out` parameter is specified, the result is a new tensor.
+
+## Example
+
+This example demonstrates how to use the `.sign()` method on a tensor containing positive, negative, and zero values:
+
+```py
+import torch
+
+# Create a tensor with mixed values
+x = torch.tensor([1.5, -2.3, 0.0, 3.7, -0.5])
+
+# Compute the sign of each element
+y = torch.sign(x)
+
+print(f"Original tensor: {x}")
+print(f"Sign tensor: {y}")
+```
+
+This example results in the following output:
+
+```shell
+Original tensor: tensor([ 1.5000, -2.3000,  0.0000,  3.7000, -0.5000])
+Sign tensor: tensor([ 1., -1.,  0.,  1., -1.])
+```
+
+In this example:
+
+- Positive values (1.5, 3.7) return 1.
+- Negative values (-2.3, -0.5) return -1.
+- Zero (0.0) returns 0.