paper-with-me

Papers

Benchmarking Distributional Alignment of Large Language Models

2024-11-08 · Nicole Meister, Carlos Guestrin, Tatsunori Hashimoto

Language models (LMs) are increasingly used as simulacra for people, yet their ability to match the distribution of views of a specific demographic group and be \textit{distributionally aligned} remains uncertain. This notion of distributional alignment is complex, as there is significant variation in the types of attributes that are simulated. Prior works have underexplored the role of three critical variables -- the question domain, steering method, and distribution expression method -- which motivates our contribution of a benchmark explicitly addressing these dimensions. We construct a dataset expanding beyond political values, create human baselines for this task, and evaluate the extent to which an LM can align with a particular group's opinion distribution to inform design choices of such simulation systems. Our analysis reveals open problems regarding if, and how, LMs can be used to simulate humans, and that LLMs can more accurately describe the opinion distribution than simulate such distributions.

📄 PDF Abstract BibTeX arXiv:2411.05403

Code (1)

nicolemeister/benchmarking-distributional-alignment 공식 구현

Tasks

Benchmarking

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

Alignment and Safety in Large Language Models: Safety Mechanisms, Training Paradigms, and Emerging Challenges

2025-07-25 · Haoran Lu, Luyang Fang, Ruidong Zhang, Xinliang Li 외 arxiv

Due to the remarkable capabilities and growing impact of large language models (LLMs), they have been deeply integrated into many aspects of society. Thus, ensuring their alignment with human values and intentions has em…

SVLTA: Benchmarking Vision-Language Temporal Alignment via Synthetic Video Situation

2025-01-01 · CVPR 2025 1 · Hao Du, Bo Wu, Yan Lu, Zhendong Mao

Vision-language temporal alignment is a crucial capability for human dynamic recognition and cognition in real-world scenarios. While existing research focuses on capturing vision-language relevance, it faces limitat…

BenchmarkingDiagnosticQuestion Answering

Quantifying Phonosemantic Iconicity Distributionally in 6 Languages

2025-10-15 · George Flint, Kaustubh Kislay arxiv

Language is, as commonly theorized, largely arbitrary. Yet, systematic relationships between phonetics and semantics have been observed in many specific cases. To what degree could those systematic relationships manifest…

Semantic Similarity

Distributional Vision-Language Alignment by Cauchy-Schwarz Divergence

2025-02-24 · Wenzhe Yin, Zehao Xiao, Pan Zhou, Shujian Yu 외

Multimodal alignment is crucial for various downstream tasks such as cross-modal generation and retrieval. Previous multimodal approaches like CLIP utilize InfoNCE to maximize mutual information, primarily aligning pairw…

Image GenerationRetrievalText to Image GenerationText-to-Image Generation

Beyond Facts: Benchmarking Distributional Reading Comprehension in Large Language Models

2026-03-13 · Pei-Fu Guo, Ya-An Tsai, Chun-Chia Hsu, Kai-Xin Chen 외 arxiv

While most reading comprehension benchmarks for LLMs focus on factual information that can be answered by localizing specific textual evidence, many real-world tasks require understanding distributional information, such…

Reading Comprehension