Presents a deterministic minimax-optimal multicalibration algorithm and its generalization to outcome indistinguishability and omniprediction, resolving open questions on randomization necessity.
Breaking the T\^ (2/3) barrier for sequential calibration
4 Pith papers cite this work. Polarity classification is still indexing.
4
Pith papers citing it
years
2026 4verdicts
UNVERDICTED 4representative citing papers
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.
Presents polynomial-time algorithms for 2D forecasting with Õ(√(kT)) swap regret and extensions to higher dimensions with Õ(d√(kT)) bounds, improving prior regret and runtime results.
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
-
Optimal Deterministic Multicalibration and Omniprediction
Presents a deterministic minimax-optimal multicalibration algorithm and its generalization to outcome indistinguishability and omniprediction, resolving open questions on randomization necessity.
-
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.
-
Improved Multi-Dimensional Forecasting for Swap Regret
Presents polynomial-time algorithms for 2D forecasting with Õ(√(kT)) swap regret and extensions to higher dimensions with Õ(d√(kT)) bounds, improving prior regret and runtime results.
-
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.