15 papers
JobBench: Aligning Agent Work With Human Will
Yuetai Li, Yichen Feng, Zhangchen Xu +21
Current benchmarks for occupational AI agents are scoped primarily by economic values, telling a replacement story. We introduce JobBench, which evaluates AI agents on the workflow…
ICONS: Influence Consensus for Vision-Language Data Selection
Xindi Wu, Mengzhou Xia, Rulin Shao +3
Training vision-language models via instruction tuning relies on large data mixtures spanning diverse tasks and domains, yet these mixtures frequently include redundant information…
Multilingual Diversity Improves Vision-Language Representations
Thao Nguyen, Matthew Wallingford, Sebastin Santy +5
Massive web-crawled image-text datasets lay the foundation for recent progress in multimodal learning. These datasets are designed with the goal of training a model to do well on s…
Fluid Language Model Benchmarking
Valentin Hofmann, David Heineman, Ian Magnusson +7
Language model (LM) benchmarking faces several challenges: comprehensive evaluations are costly, benchmarks often fail to measure the intended capabilities, and evaluation quality…
FlexOlmo: Open Language Models for Flexible Data Use
Weijia Shi, Akshita Bhagia, Kevin Farhat +20
We introduce FlexOlmo, a new class of language models (LMs) that supports (1) distributed training without data sharing, where different model parameters are independently trained…
Establishing Task Scaling Laws via Compute-Efficient Model Ladders
Akshita Bhagia, Jiacheng Liu, Alexander Wettig +9
We develop task scaling laws and model ladders to predict the individual task performance of pretrained language models (LMs) in the overtrained setting. Standard power laws for la…