There was a major disagreement between Epicurus and Aristotle/Plato as to the role of "reason" and "logic" in the determination of "truth."
This is described clearly in the following excerpt written by Phillip DeLacy in his book "Philodemus - On Methods of Inference." I have underlined in red the critical parts, in which DeLacy concludes that Aristotle held that "knowledge can be obtained only through a combination of observation and reason" and "inferences from signs are not reliable except in cases where the inferences may be converted into valid syllogisms."
Compare that to the position of Epicurus as stated by Diogenes Laertius, in which "reason" is not a part of the "Canon of truth," and "it is the reality of separate perceptions that guarantees the truth of our senses," as well as "For all our notions are derived from perceptions, either by actual contact or by analogy, or resemblance, or composition, with some slight aid from reasoning."
You can decide for yourself whether you agree with Epicurus and how to apply this distinction, but this difference in viewpoint has huge implications in seeing how far Epicurus diverged from the other Greek philosophers.
References:
Here is the appendix of Philodemus: On Methods of Inference by Phillip and Estelle De Lacy
Here is a direct link to its location in the PDF. The basic point De Lacy is making is that Aristotle (and even more strongly, Plato) held that "truth" can only be established "where the inferences may be converted into valid syllogisms." This leads to the conclusion that "knowledge can be obtained only through a combination of observation and reason.
But first before reading DeLacy remember the position of Epicurus, as stated by Diogenes Laertius, here in the RD Hicks translation, that "reason" is not a part of the canon of truth:
Now here is the De Lacy commentary on Aristotle, showing that reason/logic is essentially the ULTIMATE test of "truth":
