Stanford to host one-day workshop on recent advances in Regularity Lemmas and multicalibration
Aaroth · x · 2026-10-24
A one-day workshop on "Recent advances in Regularity Lemmas (with and without calibration)" will be held at Stanford (Gates 403) on Tuesday, November 17.
- Regularity lemmas, originating with Szemerédi (1975) and Frieze–Kannan (1999), entered complexity theory via Trevisan, Tulsiani, and Vadhan (2009), and have seen a surge of work since connecting to multiaccuracy and multicalibration in algorithmic fairness.
- Talks: Princewill Okoroafor (Harvard) on sample complexity of multicalibration with an ERM oracle; Georgy Noarov (IAS) on defensive boosting for online probabilistic forecasting; Pranay Tankala (Harvard) on agnostic learning by imitating; Konstantinos Stavropoulos (IAS) on learning under contamination and distribution shift via multiaccuracy.
- The program covers connections across complexity theory, cryptography, learning theory, information theory, and communication complexity.
More from Research
- Reuse, Don't Recompute: A 4 A.M. ICLR Story Meets Meta² Agent Evolution — furongh · 2026-10-24
- Mapping Economic Theory With Agentic AI: LLM Agents Turn Theory Papers Into Graphs — mattbeane · 2026-10-24
- Robotics model leaderboards are premature, argues practitioner: task-specific evals first — paigeinsf · 2026-10-24
- UTokyo/RIKEN lab brings CAROT cross-lingual alignment method to EMNLP, +11.2 accuracy points — Bollegala · 2026-10-24
- GPT-6 Astra solves 49 of 50 research-level physics problems, catches experts' own errors — MoonL88537 · 2026-10-24
- Rice & NVIDIA's Sensorimotor Alignment predicts robot policy success without training (r=0.717) — CSProfKGD · 2026-10-24