zeroth-order

Tag

Cards List
#zeroth-order

A Zeroth-Order Paradigm for LLM Preference Alignment

Hugging Face Daily Papers ↗ · 2026-09-16 Cached

The paper proposes Comparison-based Preference Optimization (ComPO), a zeroth-order alignment method for LLMs that uses comparison oracles to avoid likelihood displacement. It includes theoretical guarantees and experimental improvements over existing methods.

0 favorites 0 likes
#zeroth-order

Overcoming the Weakest-Link Effect in LLM-Driven Program Optimization via Heterogeneous Edit Recombination

arXiv cs.LG ↗ · 2026-08-03 Cached

Introduces HERO, an LLM-based program optimizer that overcomes the weakest-link effect by generating and recombining heterogeneous atomic edits, achieving faster convergence and higher scores across algorithmic, game, agentic, and robotic domains.

0 favorites 0 likes
#zeroth-order

A Zeroth-Order Deep Learning Method for Fully Nonlinear Parabolic Partial Differential Equations with Unknown Coefficients

arXiv cs.LG ↗ · 2026-06-25 Cached

This paper introduces a model-free deep learning method for solving high-dimensional nonlinear partial differential equations with unknown coefficients, using zeroth-order derivative estimators derived from perturbed Monte Carlo trajectories. The approach avoids automatic differentiation, provides theoretical error bounds, and demonstrates competitive performance in numerical experiments.

0 favorites 0 likes
#zeroth-order

Dominant-Layer ZO: A Single Layer Dominates Zeroth-Order Fine-Tuning of LLMs

arXiv cs.LG ↗ · 2026-06-05 Cached

This paper reveals that zeroth-order fine-tuning of LLMs is dominated by a single decoding layer, which can be identified by activation outliers, and fine-tuning only that layer matches or exceeds full-model fine-tuning with up to 4.52x speedup.

0 favorites 0 likes
← Back to home

Submit Feedback