Some docs fixes

src/Libraries/Microsoft.Extensions.AI.Abstractions/ChatCompletion/ChatMessage.cs

@@ -21,15 +21,15 @@ public ChatMessage()
     }
 
     /// <summary>Initializes a new instance of the <see cref="ChatMessage"/> class.</summary>
-    /// <param name="role">Role of the author of the message.</param>
-    /// <param name="content">Content of the message.</param>
+    /// <param name="role">The role of the author of the message.</param>
+    /// <param name="content">The contents of the message.</param>
     public ChatMessage(ChatRole role, string? content)
         : this(role, content is null ? [] : [new TextContent(content)])
     {
     }
 
     /// <summary>Initializes a new instance of the <see cref="ChatMessage"/> class.</summary>
-    /// <param name="role">Role of the author of the message.</param>
+    /// <param name="role">The role of the author of the message.</param>
     /// <param name="contents">The contents for this message.</param>
     public ChatMessage(
         ChatRole role,

src/Libraries/Microsoft.Extensions.AI.Abstractions/ChatCompletion/ChatOptions.cs

@@ -10,21 +10,46 @@ namespace Microsoft.Extensions.AI;
 public class ChatOptions
 {
     /// <summary>Gets or sets the temperature for generating chat responses.</summary>
+    /// <remarks>
+    /// This value controls the randomness of predictions made by the model. Use a lower value to decrease randomness in the response.
+    /// </remarks>
     public float? Temperature { get; set; }
 
     /// <summary>Gets or sets the maximum number of tokens in the generated chat response.</summary>
     public int? MaxOutputTokens { get; set; }
 
     /// <summary>Gets or sets the "nucleus sampling" factor (or "top p") for generating chat responses.</summary>
+    /// <remarks>
+    /// Nucleus sampling is an alternative to sampling with temperature where the model
+    /// considers the results of the tokens with <see cref="TopP"/> probability mass.
+    /// For example, 0.1 means only the tokens comprising the top 10% probability mass are considered.
+    /// </remarks>
     public float? TopP { get; set; }
 
-    /// <summary>Gets or sets a count indicating how many of the most probable tokens the model should consider when generating the next part of the text.</summary>
+    /// <summary>
+    /// Gets or sets the number of most probable tokens that the model considers when generating the next part of the text.
+    /// </summary>
+    /// <remarks>
+    /// This property reduces the probability of generating nonsense. A higher value gives more diverse answers, while a lower value is more conservative.
+    /// </remarks>
     public int? TopK { get; set; }
 
-    /// <summary>Gets or sets the frequency penalty for generating chat responses.</summary>
+    /// <summary>
+    /// Gets or sets the penalty for repeated tokens in chat responses proportional to how many times they've appeared.
+    /// </summary>
+    /// <remarks>
+    /// You can modify this value to reduce the repetitiveness of generated tokens. The higher the value, the stronger a penalty
+    /// is applied to previously present tokens, proportional to how many times they've already appeared in the prompt or prior generation.
+    /// </remarks>
     public float? FrequencyPenalty { get; set; }
 
-    /// <summary>Gets or sets the presence penalty for generating chat responses.</summary>
+    /// <summary>
+    /// Gets or sets a value that influences the probability of generated tokens appearing based on their existing presence in generated text.
+    /// </summary>
+    /// <remarks>
+    /// You can modify this value to reduce repetitiveness of generated tokens. Similar to <see cref="FrequencyPenalty"/>,
+    /// except that this penalty is applied equally to all tokens that have already appeared, regardless of their exact frequencies.
+    /// </remarks>
     public float? PresencePenalty { get; set; }
 
     /// <summary>Gets or sets a seed value used by a service to control the reproducibility of results.</summary>
@@ -47,7 +72,12 @@ public class ChatOptions
     /// <summary>Gets or sets the model ID for the chat request.</summary>
     public string? ModelId { get; set; }
 
-    /// <summary>Gets or sets the stop sequences for generating chat responses.</summary>
+    /// <summary>
+    /// Gets or sets the list of stop sequences.
+    /// </summary>
+    /// <remarks>
+    /// After a stop sequence is detected, the model stops generating further tokens for chat responses.
+    /// </remarks>
     public IList<string>? StopSequences { get; set; }
 
     /// <summary>Gets or sets the tool mode for the chat request.</summary>

src/Libraries/Microsoft.Extensions.AI/ChatCompletion/AnonymousDelegatingChatClient.cs

@@ -13,7 +13,7 @@
 
 namespace Microsoft.Extensions.AI;
 
-/// <summary>A delegating chat client that wraps an inner client with implementations provided by delegates.</summary>
+/// <summary>Represents a delegating chat client that wraps an inner client with implementations provided by delegates.</summary>
 public sealed class AnonymousDelegatingChatClient : DelegatingChatClient
 {
     /// <summary>The delegate to use as the implementation of <see cref="CompleteAsync"/>.</summary>
@@ -40,7 +40,7 @@ public sealed class AnonymousDelegatingChatClient : DelegatingChatClient
     /// used to perform the operation on the inner client. It will handle both the non-streaming and streaming cases.
     /// </param>
     /// <remarks>
-    /// This overload may be used when the anonymous implementation needs to provide pre- and/or post-processing, but doesn't
+    /// This overload may be used when the anonymous implementation needs to provide pre-processing and/or post-processing, but doesn't
     /// need to interact with the results of the operation, which will come from the inner client.
     /// </remarks>
     /// <exception cref="ArgumentNullException"><paramref name="innerClient"/> is <see langword="null"/>.</exception>

src/Libraries/Microsoft.Extensions.AI/ChatCompletion/CachingChatClient.cs

@@ -13,7 +13,7 @@
 namespace Microsoft.Extensions.AI;
 
 /// <summary>
-/// A delegating chat client that caches the results of chat calls.
+/// Represents a delegating chat client that caches the results of chat calls.
 /// </summary>
 public abstract class CachingChatClient : DelegatingChatClient
 {
@@ -30,18 +30,18 @@ protected CachingChatClient(IChatClient innerClient)
     {
     }
 
-    /// <summary>Gets or sets a value indicating whether to coalesce streaming updates.</summary>
-    /// <remarks>
+    /// <summary>Gets or sets a value indicating whether streaming updates are coalesced.</summary>
+    /// <value>
     /// <para>
-    /// When <see langword="true"/>, the client will attempt to coalesce contiguous streaming updates
-    /// into a single update, in order to reduce the number of individual items that are yielded on
-    /// subsequent enumerations of the cached data. When <see langword="false"/>, the updates are
+    /// <see langword="true"/> if the client attempts to coalesce contiguous streaming updates
+    /// into a single update, to reduce the number of individual items that are yielded on
+    /// subsequent enumerations of the cached data; <see langword="false"/> if the updates are
     /// kept unaltered.
     /// </para>
     /// <para>
     /// The default is <see langword="true"/>.
     /// </para>
-    /// </remarks>
+    /// </value>
     public bool CoalesceStreamingUpdates { get; set; } = true;
 
     /// <inheritdoc />