diff --git a/README.md b/README.md
index baefc6f..d1ce12c 100644
--- a/README.md
+++ b/README.md
@@ -19,7 +19,7 @@
 
 ### 1. 算子：SiLU函数（10分）
 
-请在`src/operators.rs`中实现SiLU算子，其公式为：
+请在`src/operators.rs`中实现SwiGLU算子，其公式为：
 
 $$
 y=silu(x) × y
@@ -85,8 +85,8 @@ $$
 hidden = rms_norm(residual)
 gate = hidden @ gate_weight.T
 up = hidden @ up_weight.T
-itermediate = gate * sigmoid(gate) * up ## silu
-output = itermediate @ down_weight.T
+act = gate * sigmoid(gate) * up ## SwiGLU
+output = act @ down_weight.T
 residual = output + residual
 ```
 
@@ -149,9 +149,9 @@ V = cat(V_cache, V)
 ### 以下是你需要实现的部分
 score = Q @ K.T / sqrt(dim)
 attn = softmax(score)
-x = attn @ V
-x = x @ O_weight.T
-residual = x + residual
+attn_V = attn @ V
+out = attn_V @ O_weight.T
+residual = out + residual
 ```
 
 Self-Attention的调试是很困难的。这里推荐大家使用pytorch来辅助调试。各位可以用transformers库（使用llama模型代码）来加载模型并运行，逐层检查中间张量结果。
diff --git a/src/main.rs b/src/main.rs
index 8226fb1..75cb4b7 100644
--- a/src/main.rs
+++ b/src/main.rs
@@ -20,8 +20,8 @@ fn main() {
     let output_ids = llama.generate(
         input_ids,
         500,
-        0.9,
-        4,
+        0.8,
+        30,
         1.,
     );
     println!("{}", tokenizer.decode(&output_ids, true).unwrap());
diff --git a/src/operators.rs b/src/operators.rs
index e2b026d..2d7355e 100644
--- a/src/operators.rs
+++ b/src/operators.rs
@@ -74,9 +74,9 @@ pub fn rms_norm(y: &mut Tensor<f32>, x: &Tensor<f32>, w: &Tensor<f32>, epsilon:
     todo!("实现 rms_norm，计算前做一些必要的检查会帮助你后续调试")
 }
 
-// y = sigmoid(x) * x * y
+// y = silu(x) * y
 // hint: this is an element-wise operation
-pub fn silu(y: &mut Tensor<f32>, x: &Tensor<f32>) {
+pub fn swiglu(y: &mut Tensor<f32>, x: &Tensor<f32>) {
     // let len = y.size();
     // assert!(len == x.size());
 
@@ -176,7 +176,7 @@ pub fn random_sample(x: &Tensor<f32>, top_p: f32, top_k: u32, temperature: f32)
 fn test_silu() {
     let mut y = Tensor::<f32>::new(vec![2., 3., 4.], &vec![1, 3]);
     let x = Tensor::<f32>::new(vec![1., 2., 3.], &vec![1, 3]);
-    silu(&mut y, &x);
+    swiglu(&mut y, &x);
     assert!(y.close_to(
         &Tensor::<f32>::new(vec![1.4621172, 5.2847824, 11.43089], &vec![1, 3]),
         1e-3