Run 2
Attempts
72
56 graded
Mean score
76.9
over graded attempts
Failures
16
errors + degenerate
Tokens out
144.7k
575 model calls
Wall time
14605s
sequential
//
glm4-9b
oneshot 87.0agentic 49.4
Expression evaluator (no eval)
agentic✓ ok
48.6
score
tests1/7
deliverables100%
content checks100%
24
model calls
13.1k
tokens out
2
SLOC
1
complexity
1289s
wall
FizzBuzz
agentic✓ ok
76
score
tests3/5
deliverables100%
content checks100%
24
model calls
2.6k
tokens out
25
SLOC
7
complexity
230s
wall
JSON Pointer (RFC 6901)
agentic✓ ok
0
score
tests0/1
deliverables0%
content checks0%
24
model calls
3.1k
tokens out
5
SLOC
2
complexity
323s
wall
LRU cache
agentic✓ ok
40
score
tests0/1
deliverables100%
content checks100%
24
model calls
3.8k
tokens out
18
SLOC
0
complexity
372s
wall
Palindrome check
agentic✓ ok
100
score
tests5/5
deliverables100%
content checks100%
4
model calls
355
tokens out
3
SLOC
3
complexity
35s
wall
Temperature conversion
agentic✓ ok
40
score
tests0/1
deliverables100%
content checks100%
24
model calls
4.5k
tokens out
13
SLOC
1
complexity
368s
wall
Text statistics module
agentic✓ ok
0
score
tests0/1
deliverables0%
content checks0%
24
model calls
7.0k
tokens out
0
SLOC
0
complexity
535s
wall
Vowel count
agentic✓ ok
100
score
tests5/5
deliverables100%
content checks100%
4
model calls
405
tokens out
3
SLOC
3
complexity
30s
wall
Word ladder (shortest path)
agentic✓ ok
40
score
tests0/5
deliverables100%
content checks100%
24
model calls
3.4k
tokens out
33
SLOC
13
complexity
280s
wall
Expression evaluator (no eval)
oneshot⚠ degenerate
—
score
tests—
deliverables—
content checks—
1
model calls
252
tokens out
0
SLOC
0
complexity
30s
wall
FizzBuzz
oneshot✓ ok
100
score
tests3/3
deliverables100%
content checks100%
1
model calls
129
tokens out
12
SLOC
6
complexity
12s
wall
JSON Pointer (RFC 6901)
oneshot✓ ok
82
score
tests7/10
deliverables100%
content checks100%
1
model calls
225
tokens out
22
SLOC
12
complexity
23s
wall
LRU cache
oneshot⚠ degenerate
—
score
tests—
deliverables—
content checks—
1
model calls
153
tokens out
0
SLOC
0
complexity
14s
wall
Palindrome check
oneshot✓ ok
100
score
tests5/5
deliverables100%
content checks100%
1
model calls
57
tokens out
3
SLOC
3
complexity
6s
wall
Temperature conversion
oneshot✓ ok
100
score
tests4/4
deliverables100%
content checks100%
1
model calls
69
tokens out
4
SLOC
1
complexity
7s
wall
Text statistics module
oneshot⚠ degenerate
—
score
tests—
deliverables—
content checks—
1
model calls
242
tokens out
0
SLOC
0
complexity
17s
wall
Vowel count
oneshot✓ ok
100
score
tests5/5
deliverables100%
content checks100%
1
model calls
46
tokens out
2
SLOC
3
complexity
4s
wall
Word ladder (shortest path)
oneshot✓ ok
40
score
tests0/5
deliverables100%
content checks100%
1
model calls
185
tokens out
17
SLOC
8
complexity
14s
wall
//
llama3.2-3b
oneshot 91.4agentic 77.9
Expression evaluator (no eval)
agentic✓ ok
91.4
score
tests6/7
deliverables100%
content checks100%
24
model calls
5.9k
tokens out
2
SLOC
1
complexity
295s
wall
FizzBuzz
agentic✓ ok
100
score
tests3/3
deliverables100%
content checks100%
3
model calls
119
tokens out
8
SLOC
5
complexity
8s
wall
JSON Pointer (RFC 6901)
agentic✓ ok
82
score
tests7/10
deliverables100%
content checks100%
24
model calls
5.3k
tokens out
16
SLOC
8
complexity
305s
wall
LRU cache
agentic✓ ok
76
score
tests3/5
deliverables100%
content checks100%
24
model calls
2.0k
tokens out
19
SLOC
4
complexity
111s
wall
Palindrome check
agentic✓ ok
100
score
tests5/5
deliverables100%
content checks100%
3
model calls
77
tokens out
3
SLOC
3
complexity
5s
wall
Temperature conversion
agentic✓ ok
100
score
tests4/4
deliverables100%
content checks100%
3
model calls
91
tokens out
4
SLOC
1
complexity
5s
wall
Text statistics module
agentic✓ ok
0
score
tests0/1
deliverables0%
content checks0%
24
model calls
1.2k
tokens out
0
SLOC
1
complexity
67s
wall
Vowel count
agentic✓ ok
100
score
tests5/5
deliverables100%
content checks100%
3
model calls
75
tokens out
3
SLOC
3
complexity
4s
wall
Word ladder (shortest path)
agentic✓ ok
52
score
tests1/5
deliverables100%
content checks100%
24
model calls
1.6k
tokens out
17
SLOC
9
complexity
119s
wall
Expression evaluator (no eval)
oneshot✓ ok
40
score
tests0/1
deliverables100%
content checks100%
1
model calls
254
tokens out
18
SLOC
4
complexity
12s
wall
FizzBuzz
oneshot✓ ok
100
score
tests3/3
deliverables100%
content checks100%
1
model calls
93
tokens out
8
SLOC
5
complexity
4s
wall
JSON Pointer (RFC 6901)
oneshot⚠ degenerate
—
score
tests—
deliverables—
content checks—
1
model calls
171
tokens out
0
SLOC
0
complexity
7s
wall
LRU cache
oneshot✓ ok
100
score
tests5/5
deliverables100%
content checks100%
1
model calls
157
tokens out
18
SLOC
4
complexity
6s
wall
Palindrome check
oneshot✓ ok
100
score
tests5/5
deliverables100%
content checks100%
1
model calls
52
tokens out
3
SLOC
3
complexity
3s
wall
Temperature conversion
oneshot✓ ok
100
score
tests4/4
deliverables100%
content checks100%
1
model calls
68
tokens out
5
SLOC
1
complexity
3s
wall
Text statistics module
oneshot⚠ degenerate
—
score
tests—
deliverables—
content checks—
1
model calls
155
tokens out
0
SLOC
0
complexity
6s
wall
Vowel count
oneshot✓ ok
100
score
tests5/5
deliverables100%
content checks100%
1
model calls
43
tokens out
2
SLOC
3
complexity
2s
wall
Word ladder (shortest path)
oneshot✓ ok
100
score
tests5/5
deliverables100%
content checks100%
1
model calls
196
tokens out
19
SLOC
9
complexity
9s
wall
//
qwen2.5-coder-3b
oneshot 87.2agentic 49.8
Expression evaluator (no eval)
agentic✓ ok
48.6
score
tests1/7
deliverables100%
content checks100%
24
model calls
9.4k
tokens out
2
SLOC
1
complexity
341s
wall
FizzBuzz
agentic✓ ok
100
score
tests3/3
deliverables100%
content checks100%
24
model calls
3.3k
tokens out
12
SLOC
6
complexity
129s
wall
JSON Pointer (RFC 6901)
agentic✓ ok
0
score
tests0/1
deliverables0%
content checks0%
24
model calls
3.5k
tokens out
0
SLOC
0
complexity
139s
wall
LRU cache
agentic✓ ok
0
score
tests0/1
deliverables0%
content checks0%
24
model calls
5.5k
tokens out
0
SLOC
0
complexity
239s
wall
Palindrome check
agentic✓ ok
100
score
tests5/5
deliverables100%
content checks100%
24
model calls
1.5k
tokens out
3
SLOC
3
complexity
63s
wall
Temperature conversion
agentic✓ ok
100
score
tests4/4
deliverables100%
content checks100%
24
model calls
1.9k
tokens out
4
SLOC
1
complexity
77s
wall
Text statistics module
agentic✓ ok
0
score
tests0/1
deliverables0%
content checks0%
24
model calls
8.1k
tokens out
0
SLOC
0
complexity
384s
wall
Vowel count
agentic✓ ok
100
score
tests5/5
deliverables100%
content checks100%
24
model calls
1.4k
tokens out
2
SLOC
3
complexity
63s
wall
Word ladder (shortest path)
agentic✓ ok
0
score
tests0/1
deliverables0%
content checks0%
24
model calls
5.1k
tokens out
0
SLOC
0
complexity
228s
wall
Expression evaluator (no eval)
oneshot✓ ok
48.6
score
tests1/7
deliverables100%
content checks100%
1
model calls
500
tokens out
47
SLOC
23
complexity
16s
wall
FizzBuzz
oneshot✓ ok
100
score
tests3/3
deliverables100%
content checks100%
1
model calls
138
tokens out
12
SLOC
6
complexity
6s
wall
JSON Pointer (RFC 6901)
oneshot⚠ degenerate
—
score
tests—
deliverables—
content checks—
1
model calls
171
tokens out
0
SLOC
0
complexity
6s
wall
LRU cache
oneshot⚠ degenerate
—
score
tests—
deliverables—
content checks—
1
model calls
148
tokens out
0
SLOC
0
complexity
6s
wall
Palindrome check
oneshot⚠ degenerate
—
score
tests—
deliverables—
content checks—
1
model calls
89
tokens out
0
SLOC
0
complexity
3s
wall
Temperature conversion
oneshot✓ ok
100
score
tests4/4
deliverables100%
content checks100%
1
model calls
79
tokens out
4
SLOC
1
complexity
5s
wall
Text statistics module
oneshot⚠ degenerate
—
score
tests—
deliverables—
content checks—
1
model calls
303
tokens out
0
SLOC
0
complexity
14s
wall
Vowel count
oneshot✓ ok
100
score
tests5/5
deliverables100%
content checks100%
1
model calls
60
tokens out
3
SLOC
3
complexity
3s
wall
Word ladder (shortest path)
oneshot⚠ degenerate
—
score
tests—
deliverables—
content checks—
1
model calls
250
tokens out
0
SLOC
0
complexity
12s
wall
//
qwen3-30b-a3b
oneshot 100.0agentic 100.0
Expression evaluator (no eval)
agentic✗ error
—
score
tests—
deliverables—
content checks—
0
model calls
0
tokens out
0
SLOC
0
complexity
906s
wall
FizzBuzz
agentic✓ ok
100
score
tests3/3
deliverables100%
content checks100%
3
model calls
4.5k
tokens out
12
SLOC
6
complexity
215s
wall
JSON Pointer (RFC 6901)
agentic✗ error
—
score
tests—
deliverables—
content checks—
2
model calls
10.3k
tokens out
26
SLOC
11
complexity
1400s
wall
LRU cache
agentic✓ ok
100
score
tests5/5
deliverables100%
content checks100%
3
model calls
3.5k
tokens out
18
SLOC
4
complexity
336s
wall
Palindrome check
agentic✓ ok
100
score
tests5/5
deliverables100%
content checks100%
3
model calls
5.5k
tokens out
3
SLOC
3
complexity
380s
wall
Temperature conversion
agentic✓ ok
100
score
tests4/4
deliverables100%
content checks100%
3
model calls
2.1k
tokens out
4
SLOC
1
complexity
132s
wall
Text statistics module
agentic✗ error
—
score
tests—
deliverables—
content checks—
0
model calls
0
tokens out
0
SLOC
0
complexity
906s
wall
Vowel count
agentic✓ ok
100
score
tests5/5
deliverables100%
content checks100%
3
model calls
2.0k
tokens out
3
SLOC
3
complexity
107s
wall
Word ladder (shortest path)
agentic✗ error
—
score
tests—
deliverables—
content checks—
0
model calls
0
tokens out
0
SLOC
0
complexity
906s
wall
Expression evaluator (no eval)
oneshot✗ error
—
score
tests—
deliverables—
content checks—
0
model calls
0
tokens out
0
SLOC
0
complexity
906s
wall
FizzBuzz
oneshot✓ ok
100
score
tests3/3
deliverables100%
content checks100%
1
model calls
1.8k
tokens out
2
SLOC
5
complexity
89s
wall
JSON Pointer (RFC 6901)
oneshot✓ ok
100
score
tests10/10
deliverables100%
content checks100%
1
model calls
5.4k
tokens out
23
SLOC
8
complexity
284s
wall
LRU cache
oneshot✓ ok
100
score
tests5/5
deliverables100%
content checks100%
1
model calls
1.8k
tokens out
18
SLOC
4
complexity
238s
wall
Palindrome check
oneshot✓ ok
100
score
tests5/5
deliverables100%
content checks100%
1
model calls
2.8k
tokens out
3
SLOC
1
complexity
291s
wall
Temperature conversion
oneshot✓ ok
100
score
tests4/4
deliverables100%
content checks100%
1
model calls
838
tokens out
4
SLOC
1
complexity
52s
wall
Text statistics module
oneshot✗ error
—
score
tests—
deliverables—
content checks—
0
model calls
0
tokens out
0
SLOC
0
complexity
906s
wall
Vowel count
oneshot✓ ok
100
score
tests5/5
deliverables100%
content checks100%
1
model calls
970
tokens out
2
SLOC
3
complexity
60s
wall
Word ladder (shortest path)
oneshot✓ ok
100
score
tests5/5
deliverables100%
content checks100%
1
model calls
4.4k
tokens out
21
SLOC
9
complexity
196s
wall