fix: normalize double-escaped LaTeX backslashes and HTML entities (#14564)

Fixes #14562

## Problem
LLMs like DeepSeek V4 Flash and Qwen3-MAX return \\( and \\[ 
(double backslash) in LaTeX output. The preprocessLaTeX() function 
only handled single backslash delimiters, so equations showed as raw
text.
HTML entities like < and > were also not decoded.

## Solution
Added normalization step before existing delimiter conversion:
- \\( → \(  and  \\[ → \[
- &lt; → <  and  &gt; → >  and  &amp; → &

---------

Co-authored-by: Vivek <viveksantoshkumardubey@email.com>

web/src/utils/tests/chat.test.ts

@@ -0,0 +1,26 @@
+import { preprocessLaTeX } from '../chat';
+
+test('handles double-escaped inline LaTeX', () => {
+  const result = preprocessLaTeX('\\\\(\\\\Delta = b^2\\\\)');
+  expect(result).toBe('$\\Delta = b^2$');
+});
+
+test('handles double-escaped block LaTeX', () => {
+  const result = preprocessLaTeX('\\\\[E = mc^2\\\\]');
+  expect(result).toBe('$$E = mc^2$$');
+});
+
+test('decodes HTML entities', () => {
+  const result = preprocessLaTeX('a < b & c > d');
+  expect(result).toBe('a < b & c > d');
+});
+
+test('handles mixed double-escaped delimiters with HTML entities', () => {
+  const result = preprocessLaTeX('\\\\(x &lt; y\\\\)');
+  expect(result).toBe('$x < y$');
+});
+
+test('passes through already correct single-escaped delimiters unchanged', () => {
+  const result = preprocessLaTeX('\\(x = 1\\)');
+  expect(result).toBe('$x = 1$');
+});