AlphaPO: Reward Shape Matters for LLM Alignment

Documenti analoghi

• On the Robustness of Reward Models for Language Model Alignment
  di: Hong, Jiwoo, et al.
  Pubblicazione: (2025)
• FFSplit: Split Feed-Forward Network For Optimizing Accuracy-Efficiency Trade-off in Language Model Inference
  di: Liu, Zirui, et al.
  Pubblicazione: (2024)
• Microergodicity implies orthogonality of Matérn fields on bounded domains in $\mathbb{R}^4$
  di: Pillai, Natesh S.
  Pubblicazione: (2026)
• Optimal Scaling for the Proximal Langevin Algorithm in High Dimensions
  di: Pillai, Natesh S.
  Pubblicazione: (2022)
• A Precise Characterization of SGD Stability Using Loss Surface Geometry
  di: Dexter, Gregory, et al.
  Pubblicazione: (2024)