We evaluate polyfuzz on several popular open-source projects developed mainly in 10 Python-C programs, 5 Java-C programs, and 15 single-language programs (5 respectively in C, Python, and Java). Eventually, 14 vulnerabilities in 8 projects below are validated to be exploitable, and corresponding PoCs are attached.

Affected version: version <= 5.2.0
Description: With carefully constructed inputs, NULL pointer reference can happen and cause segment fault. This allows attackers to conduct DoS attacks by construct specific inputs.
Exploitation: PoC: ujson_poc.py with Input and Output.

Affected version: version <= 6.0
Description: With carefully constructed inputs, Recursion Error can happen. This allows attackers to conduct DoS attacks by construct specific inputs.
Exploitation: PoC: poc_load.py with Input and Output.

Affected version: version <= 2.4.0
Description: With carefully constructed inputs, Out of Memory can happen. This allows attackers to conduct DoS attacks by construct specific inputs.
Exploitation: PoC: OutStream with Input.

Affected version: version <= 9.1.1
Description: With carefully constructed inputs, out of memory can happen in API convert. This allows attackers to conduct DoS attacks by construct specific inputs
Exploitation: PoC: poc_fig_process.py with Input.

Affected version: version <= 2.4.3
Description: With carefully constructed inputs, libsmbios can crash with bus error. This allows attackers to conduct DoS attacks by construct specific inputs
Exploitation: PoC: poc_op_mem.py with Input and Output.

Affected version: version <= 3.24.2
Description: With carefully constructed inputs, JVM hangs. This allows attackers to conduct DoS attacks by construct specific inputs
Exploitation: PoC: jparser with Input and Output.

Affected version: version <= 0.4.9
Description: With carefully constructed inputs, memory leak could happen during continuous running.
Exploitation: PoC: filter-test.py with Input and Output.

Affected version: version <= 1.3.4
Description: With carefully constructed inputs, the API median can crash with segment fault. This allows attackers to conduct DoS attacks by construct specific inputs.
Exploitation: PoC: random_shape.py and Output.
With 7 inputs: input1, input2, input3, input4, input5, input6, input7

Affected version: version <= 1.3.4
Description: With carefully constructed inputs, the API nanmean can crash with segment fault. This allows attackers to conduct DoS attacks by construct specific inputs.
Exploitation: PoC: random_shape.py and Output.
With 7 inputs: input1, input2, input3, input4, input5, input6, input7

Affected version: version <= 1.3.4
Description: With carefully constructed inputs, the API nanmedian can crash with segment fault. This allows attackers to conduct DoS attacks by construct specific inputs.
Exploitation: PoC: random_shape.py and Output.
With 7 inputs: input1, input2, input3, input4, input5, input6, input7

Affected version: version <= 1.3.4
Description: With carefully constructed inputs, the API nanmin can crash with segment fault. This allows attackers to conduct DoS attacks by construct specific inputs.
Exploitation: PoC: random_shape.py and Output.
With 7 inputs: input1, input2, input3, input4, input5, input6, input7

Affected version: version <= 1.3.4
Description: With carefully constructed inputs, the API nanstd can crash with segment fault. This allows attackers to conduct DoS attacks by construct specific inputs.
Exploitation: PoC: random_shape.py and Output.
With 7 inputs: input1, input2, input3, input4, input5, input6, input7

Affected version: version <= 1.3.4
Description: With carefully constructed inputs, the API ss can crash with segment fault. This allows attackers to conduct DoS attacks by construct specific inputs.
Exploitation: PoC: random_shape.py and Output.
With 7 inputs: input1, input2, input3, input4, input5, input6, input7

Affected version: version <= 1.3.4
Description: With carefully constructed inputs, the API nanmax can crash with segment fault. This allows attackers to conduct DoS attacks by construct specific inputs.
Exploitation: PoC: random_shape.py and Output.
With 7 inputs: input1, input2, input3, input4, input5, input6, input7