paper-with-me

Papers

Deliberative Alignment: Reasoning Enables Safer Language Models

2024-12-20 · Melody Y. Guan, Manas Joglekar, Eric Wallace, Saachi Jain, Boaz Barak, Alec Helyar, Rachel Dias, Andrea Vallone, Hongyu Ren, Jason Wei, Hyung Won Chung, Sam Toyer, Johannes Heidecke, Alex Beutel, Amelia Glaese

As large-scale language models increasingly impact safety-critical domains, ensuring their reliable adherence to well-defined principles remains a fundamental challenge. We introduce Deliberative Alignment, a new paradigm that directly teaches the model safety specifications and trains it to explicitly recall and accurately reason over the specifications before answering. We used this approach to align OpenAI's o-series models, and achieved highly precise adherence to OpenAI's safety policies, without requiring human-written chain-of-thoughts or answers. Deliberative Alignment pushes the Pareto frontier by simultaneously increasing robustness to jailbreaks while decreasing overrefusal rates, and also improves out-of-distribution generalization. We demonstrate that reasoning over explicitly specified policies enables more scalable, trustworthy, and interpretable alignment.

📄 PDF Abstract BibTeX arXiv:2412.16339

Code (0)

등록된 구현이 없습니다.

Tasks

Out-of-Distribution Generalization

Methods 이 논문이 사용한 방법론

ALIGN In the ALIGN method, visual and language representations are jointly trained from noisy image alt-text data. The image and text encoders are learned via contrastive loss…

Similar Papers 제목 키워드 기반

STAR-1: Safer Alignment of Reasoning LLMs with 1K Data

2025-04-02 · Zijun Wang, Haoqin Tu, YuHan Wang, Juncheng Wu 외

This paper introduces STAR-1, a high-quality, just-1k-scale safety dataset specifically designed for large reasoning models (LRMs) like DeepSeek-R1. Built on three core principles -- diversity, deliberative reasoning, an…

DiversitySafety Alignment

EASE: Practical and Efficient Safety Alignment for Small Language Models

2025-11-09 · Haonan Shi, Guoli Wang, Tu Ouyang, An Wang arxiv

Small language models (SLMs) are increasingly deployed on edge devices, making their safety alignment crucial yet challenging. Current shallow alignment methods that rely on direct refusal of malicious queries fail to pr…

Computational Efficiency

Large Language Models for Recommendation with Deliberative User Preference Alignment

2025-02-04 · Yi Fang, Wenjie Wang, Yang Zhang, Fengbin Zhu 외

While recent advancements in aligning Large Language Models (LLMs) with recommendation tasks have shown great potential and promising performance overall, these aligned recommendation LLMs still face challenges in comple…

Deliberative Alignment is Deep, but Uncertainty Remains: Inference time safety improvement in reasoning via attribution of unsafe behavior to base model

2026-04-01 · Pankayaraj Pathmanathan, Furong Huang arxiv

While the wide adoption of refusal training in large language models (LLMs) has showcased improvements in model safety, recent works have highlighted shortcomings due to the shallow nature of these alignment methods. To …

Reasoning over Precedents Alongside Statutes: Case-Augmented Deliberative Alignment for LLM Safety

2026-01-12 · Can Jin, Rui Wu, Tong Che, Qixin Zhang 외 arxiv

Ensuring that Large Language Models (LLMs) adhere to safety principles without refusing benign requests remains a significant challenge. While OpenAI introduces deliberative alignment (DA) to enhance the safety of its o-…

Reinforcement Learning