The Deterministic Horizon: When Extended Reasoning Fails and Tool Delegation Becomes Necessary 事件

The Deterministic Horizon: When Extended Reasoning Fails and Tool Delegation Becomes Necessary arXiv:2606.00376v1 Announce Type: cross Abstract: Extended chain-of-thought reasoning can degrade performance on deterministic state-tracking tasks, not due to preference biases, but limits rooted in the information-theoretic capacity of decoder-only attention. We establish: (1) an Attention Bottleneck Theorem with a complementary achievability construction, bounding state-tracking capacity as $O(H \c