Currently in ML-KEM's 12-bit decompression method (twelve2sixteen()) the assertion is incorrect:

            assert ((remainder == 0) || (remainder == 48)) &&
            (index + i * 96 <= condensed.length);
    implKyber12To16(condensed, index, parsed, parsedLength);

From [~frakoczi]: "All of the calls are processing 192 or 384 bytes (and producing 128 or 256 shorts)."