Attention and LoRA regression losses induce Poincaré inequalities under mild regularization, so SGD-mimicking SDEs converge to minimizers with no assumptions on data or model size.
Breaking the T\^ (2/3) barrier for sequential calibration
2 Pith papers cite this work. Polarity classification is still indexing.
2
Pith papers citing it
fields
cs.LG 2years
2026 2verdicts
UNVERDICTED 2representative citing papers
Introduces SCDL as a calibration measure that is fully actionable for full swap regret and testable with nearly optimal sample error while satisfying continuity and consistency.
citing papers explorer
-
Convergent Stochastic Training of Attention and Understanding LoRA
Attention and LoRA regression losses induce Poincaré inequalities under mild regularization, so SGD-mimicking SDEs converge to minimizers with no assumptions on data or model size.
-
Testable and Actionable Calibration for Full Swap Regret
Introduces SCDL as a calibration measure that is fully actionable for full swap regret and testable with nearly optimal sample error while satisfying continuity and consistency.