REVIEW 8 cited by
Performative Prediction: Past and Future
Not yet reviewed by Pith; the record is open.
This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.
SPECIMEN: schema-true, not a live event
T0 review · schema-true
One-sentence machine reading of the paper's core claim.
pith:XXXXXXXX · record.json · timestamp
Signed reviews
read the original abstract
Predictions in the social world generally influence the target of prediction, a phenomenon known as performativity. Self-fulfilling and self-negating predictions are examples of performativity. Of fundamental importance to economics, finance, and the social sciences, the notion has been absent from the development of machine learning that builds on the static perspective of pattern recognition. In machine learning applications, however, performativity often surfaces as distribution shift. A predictive model deployed on a digital platform, for example, influences behavior and thereby changes the data-generating distribution. We discuss the recently founded area of performative prediction that provides a definition and conceptual framework to study performativity in machine learning. A key element of performative prediction is a natural equilibrium notion that gives rise to new optimization challenges. What emerges is a distinction between learning and steering, two mechanisms at play in performative prediction. Steering is in turn intimately related to questions of power in digital markets. The notion of performative power that we review gives an answer to the question how much a platform can steer participants through its predictions. We end on a discussion of future directions, such as the role that performativity plays in contesting algorithmic systems.
Forward citations
Cited by 8 Pith papers
-
Robustness or Crowding: Experimental Design for Trading Strategy Capacity
A same-date comparison of differently sized strategy sleeves identifies only the sleeve's own response at fixed aggregate positioning, never the strategy's total capacity under crowding, because an arbitrary calendar ...
-
Toward a Theory of Value in AI Alignment
A systematic annotation of 94 AI alignment papers shows the field largely equates human values with measurable preferences, rarely defines values, and is increasingly removing humans from alignment evaluation.
-
Robust Conformal CBF and CLF Controllers via Iterative Policy Updates
An iterative conformal-prediction update rule transfers probabilistic safety/stability guarantees across changing robust CBF/CLF policies despite policy-induced distribution shift.
-
Optimal Regularization for Performative Learning
Optimal ridge regularization in performative linear regression is set by the mean performative strength, and over-parameterization can turn a self-reinforcing performative effect into a lower optimally tuned risk.
-
Performative Risk Control: Calibrating Models for Reliable Deployment under Performativity
An iterative threshold calibration method, Performative Risk Control, gives finite-sample guarantees that risk stays controlled under performative (self-influencing) distribution shifts.
-
On Corruption-Robustness in Performative Reinforcement Learning
A repeated retraining algorithm with robust gradient estimation converges to an approximately stable policy in performative RL under Huber contamination, with approximation error scaling as the square root of the corr...
-
Independent Learning in Performative Markov Potential Games
In performative Markov potential games, independent policy gradient and natural policy gradient algorithms are claimed to converge to an approximate performatively stable equilibrium, but the proofs contain a false eq...
-
From Fairness to Infinity: Outcome-Indistinguishable (Omni)Prediction in Evolving Graphs
The Any Kernel algorithm, a randomized extension of Vovk's K29*, achieves online outcome indistinguishability for any RKHS and yields the first O(√T) online omnipredictors for infinite real-valued comparator classes.
Discussion (0). Continue with ORCID to comment.