diff --git a/twasm/.bootle/disk b/twasm/.bootle/disk
deleted file mode 100644
index a115815..0000000
Binary files a/twasm/.bootle/disk and /dev/null differ
diff --git a/twasm/README.md b/twasm/README.md
index cd2fa8b..83dab90 100644
--- a/twasm/README.md
+++ b/twasm/README.md
@@ -35,11 +35,11 @@ each word represents a token on the token table.
 each token gets loaded into the token table with the following form:
 
 ```
-+----------+----------+
-| 31    16 | 15     0 |
-+----------+----------+
-| reserved | token id |
-+----------+----------+
++----------+
+| 15     0 |
++----------+
+| token id |
++----------+
 ```
 
 ### token IDs
diff --git a/twasm/asm/main.asm b/twasm/asm/main.asm
index 2e5710c..e4d8b86 100644
--- a/twasm/asm/main.asm
+++ b/twasm/asm/main.asm
@@ -5,8 +5,8 @@ TEST_ARENA_SIZE equ 0x1000     ; maximum size tests can use
 
 TOKEN_TABLE_ADDR equ 0x00060000 ; address the token table is loaded at
 TOKEN_TABLE_SIZE equ 0x1000     ; max length of table
-TOKEN_TABLE_ENTRY_SIZE equ 8    ; size of token table entry; a LOT of things
-                                ; may break if this ever changes
+TOKEN_TABLE_ENTRY_SIZE equ 2    ; size of token table entry; things may break
+                                ; if this ever changes
 
 OUTPUT_ADDR equ 0x00070000 ; address of outputed binary
 OUTPUT_SIZE equ 0x1000     ; max length of outputed binary
@@ -84,10 +84,9 @@ tokenise:
     mov edx, 1     ; byte length is 1
 
   .continue:
-
     add rdi, rdx ; current byte + length of token = next unread byte
 
-    mov [TOKEN_TABLE_ADDR + rcx * TOKEN_TABLE_ENTRY_SIZE], rax ; fill next entry
+    mov [TOKEN_TABLE_ADDR + rcx * TOKEN_TABLE_ENTRY_SIZE], ax ; fill next entry
                                                                ; in token table
     inc rcx ; +1 token processed
     jmp .loop