Signals 4 · free daily AI digest

A Zeroth-Order Paradigm for LLM Preference Alignment

Paper recorded by Signals 4 on 2026-09-16 in cs.AI. Abstract reproduced from arXiv; link to the original below.

Published 2026-09-16 on arXiv · recorded by Signals 4 on 2026-09-17

Category: cs.AI · 人工智能 · first seen 2026-09-17

Abstract

Direct preference alignment methods are widely used to align large language models (LLMs) with human preferences because of their computational and memory efficiency. However, likelihood displacement motivates alternative ways to extract information from preference pairs with small likelihood margins. In this paper, we propose and analyze Comparison-based Preference Optimization (ComPO), a zeroth-

Read on arXiv →

#22 most recent of 300 cs.AI papers we have recorded · ↑ newer: Objective vs. Search: Decomposing What Makes a Good Tokeniser · ↓ older: Dreaming the Sound of Contact: Leveraging Video and Audio Generation f
Cite this page: A Zeroth-Order Paradigm for LLM Preference Alignment: the #22 most recent of 300 cs.AI papers we have recorded (as of 2026-09-16). Source: Signals 4 (Signals API) — https://data.jiangzhang.ca/signals4/t/papers/a-zeroth-order-paradigm-for-llm-preference-alignment.html
Free to quote with attribution to “Signals 4 (Signals API)”. Machine-readable: papers.json
Related: More cs.AI papers · arXiv signals · All papers · Today in AI
Get 4 AI signals a day by email — free.
Subscribe free → See all plans →
Get 4 AI signals a day by email — free
All models · All repos · By company · Daily editions