Skip to main content ->
Ai2

Research - Papers

Explore a selection of our published work on a variety of key research challenges in AI.

Filter papers

Understanding the Logic of Direct Preference Alignment through Logic

Kyle RichardsonVivek SrikumarAshish Sabharwal
2025
ICML

Recent direct preference alignment algorithms (DPA), such as DPO, have shown great promise in aligning large language models to human preferences. While this has motivated the development of many… 

OLMoTrace: Tracing Language Model Outputs Back to Trillions of Training Tokens

Jiacheng LiuTaylor BlantonYanai ElazarJesse Dodge
2025
Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 3: System Demonstrations)

We present OLMoTrace, the first system that traces the outputs of language models back to their full, multi-trillion-token training data in real time. OLMoTrace finds and shows verbatim matches… 

Critical Batch Size Revisited: A Simple Empirical Approach to Large-Batch Language Model Training

William MerrillShane AroraDirk GroeneveldHanna Hajishirzi
2025
arXiv.org

The right batch size is important when training language models at scale: a large batch size is necessary for fast training, but a batch size that is too large will harm token efficiency. To… 

Holodeck: Language Guided Generation of 3D Embodied AI Environments

Yue YangFan-Yun SunLuca WeihsChristopher Clark
2025
Computer Vision and Pattern Recognition

3D simulated environments play a critical role in Embodied AI, but their creation requires expertise and extensive manual effort, restricting their diversity and scope. To miti-gate this limitation,… 

What Sets the Tropical Cold Point in GSRMs During Boreal Winter? Overshooting Convection Versus Cirrus Lofting

Jacqueline M. NugentChristopher S. BrethertonP. Blossey
2025
Earth and Space Science

The cold point tropopause, the minimum temperature within the tropical upper troposphere‐lower stratosphere region (UTLS), significantly impacts Earth's climate by influencing the amount of water… 

Multi-Attribute Constraint Satisfaction via Language Model Rewriting

Ashutosh BahetiDebanjana ChakrabortyFaeze BrahmanMaarten Sap
2025
TMLR

Obeying precise constraints on top of multiple external attributes is a common computational problem underlying seemingly different domains, from controlled text generation to protein engineering.… 

ACE2: accurately learning subseasonal to decadal atmospheric variability and forced responses

Oliver Watt‐MeyerBrian HennJeremy McGibbonChristopher S. Bretherton
2025
NPJ Climate and Atmospheric Science

Existing machine learning models of weather variability are not formulated to enable assessment of their response to varying external boundary conditions such as sea surface temperature and… 

Critical Batch Size Revisited: A Simple Empirical Approach to Large-Batch Language Model Training

William MerrillShane AroraDirk GroeneveldHanna Hajishirzi
2025
arXiv.org

The right batch size is important when training language models at scale: a large batch size is necessary for fast training, but a batch size that is too large will harm token efficiency. To… 

Applying the ACE2 Emulator to SST Green's Functions for the E3SMv3 Global Atmosphere Model

Elynn WuF. RebassooPappu PaulChristopher S. Bretherton
2025
Journal of Geophysical Research - Machine Learning

Green's functions are a useful technique for interpreting atmospheric state responses to changes in the spatial pattern of sea surface temperature (SST). Here we train version 2 of the Ai2 Climate… 

RewardBench: Evaluating Reward Models for Language Modeling

Nathan LambertValentina PyatkinJacob Daniel MorrisonHanna Hajishirzi
2025
NAACL Findings

Reward models (RMs) are at the crux of successfully using RLHF to align pretrained models to human preferences, yet there has been relatively little study that focuses on evaluation of those models.…