It's interesting how clearly you've distilled this Apple report; the focus on demand stability and services growth realy reframes the narrative beyond just the numbers. I'm especially curious about the 'deep network reasoning' and 'causal graphs' mentioned in your method, it makes me wonder how these tools might eventually help us model broader public policy impacts from corporate actions.
Basically - the output is actually two arrays of numbers representing the entities and the trades. All intermediate values are computed by a deep (>10-layer) network with internal relationship mappings (there are thousands of “dots” in that "graph"). The causal graph emerges in the intermediate layers.
We do trade on those, but we also built a sub-system to “translate” the values and links embedded in the causal graph into relationships, reasoning, etc., so there’s a human-readable English version. We read it ourselves and share it with everyone so the model’s output doesn’t stay as cold numbers.
Thanks for reading and engaging — comments like yours make the research worth sharing.
It's interesting how clearly you've distilled this Apple report; the focus on demand stability and services growth realy reframes the narrative beyond just the numbers. I'm especially curious about the 'deep network reasoning' and 'causal graphs' mentioned in your method, it makes me wonder how these tools might eventually help us model broader public policy impacts from corporate actions.
Basically - the output is actually two arrays of numbers representing the entities and the trades. All intermediate values are computed by a deep (>10-layer) network with internal relationship mappings (there are thousands of “dots” in that "graph"). The causal graph emerges in the intermediate layers.
We do trade on those, but we also built a sub-system to “translate” the values and links embedded in the causal graph into relationships, reasoning, etc., so there’s a human-readable English version. We read it ourselves and share it with everyone so the model’s output doesn’t stay as cold numbers.
Thanks for reading and engaging — comments like yours make the research worth sharing.