AM
Azalia Mirhoseini
cs.LGcs.AIcs.CLstat.MLcs.DCcs.ARcs.DBcs.MAcs.PFcs.RO
On Valency
published · living versionsW_34tet3c7·v1 · currentpublished
Constitutional AI: Harmlessness from AI Feedback
with Yuntao Bai, Saurav Kadavath, Sandipan Kundu, Amanda Askell +46
1 version
Preprints & journals
58 papers in the corpus · 2015–2026Intelligence per Watt: Measuring Intelligence Efficiency of Local AI2511.07885v7 · Jon Saad-Falcon, Avanika Narayan, Hakki Orhun Akengin et al.2025 · 0 citationsarXiv
Decentralized Multi-Agent Systems with Shared Context2606.10662v2 · Yuzhen Mao, Jerry Gu, Aadi Chauhan et al.2026 · 0 citationsarXiv
Federation of Experts: Communication Efficient Distributed Inference for Large Language Models2605.06206v2 · Muhammad Shahir Abdurrahman, Chun Deng, Azalia Mirhoseini et al.2026 · 0 citationsarXiv
Block Parallelism For Efficient Distributed Long-Context Diffusion Language Model Training2609.19242v1 · Tarun Suresh, Pranshu Chaturvedi, Hangoo Kang et al.2026 · 0 citationsarXiv
KernelBench-Verified: Do LLM-Generated Kernels Actually Beat PyTorch?2607.16241v1 · Yunxiang Zhang, Ping Yu, Jianyu Wang et al.2026 · 0 citationsarXiv
LLM-as-a-Verifier: A General-Purpose Verification Framework2607.05391v2 · Jacky Kwok, Shulu Li, Pranav Atreya et al.2026 · 0 citationsarXiv
TRACE: Capability-Targeted Agentic Training2604.05336v2 · Hangoo Kang, Tarun Suresh, Jon Saad-Falcon et al.2026 · 0 citationsarXiv
Distill to Detect: Exposing Stealth Biases in LLMs through Cartridge Distillation2607.01208v1 · Shayan Talaei, Abhinav Chinta, Devvrit Khatri et al.2026 · 0 citationsarXiv
Agent JIT Compilation for Latency-Optimizing Web Agent Planning and Scheduling2605.21470v2 · Caleb Winston, Ron Yifeng Wang, Azalia Mirhoseini et al.2026 · 0 citationsarXiv
OpenJarvis: Personal AI, On Personal Devices2605.17172v1 · Jon Saad-Falcon, Avanika Narayan, Robby Manihani et al.2026 · 0 citationsarXiv
AI+HW 2035: Shaping the Next Decade2603.05225v1 · Deming Chen, Jason Cong, Azalia Mirhoseini et al.2026 · 0 citationsarXiv
Scaling Verification Can Be More Effective than Scaling Policy Learning for Vision-Language-Action Alignment2602.12281v2 · Jacky Kwok, Xilun Zhang, Mengdi Xu et al.2026 · 0 citationsarXiv
SPRINT: Enabling Interleaved Planning and Parallelized Execution in Reasoning Models2506.05745v2 · Emil Biju, Shayan Talaei, Zhemin Huang et al.2025 · 0 citationsarXiv
Astra: A Multi-Agent System for GPU Kernel Performance Optimization2509.07506v2 · Anjiang Wei, Tianran Sun, Yogesh Seenichamy et al.2025 · 0 citationsarXiv
ForTIFAI: Fending Off Recursive Training Induced Failure for AI Model Collapse2509.08972v4 · Soheil Zibakhsh Shabgahi, Pedram Aghazadeh, Azalia Mirhoseini et al.2025 · 0 citationsarXiv
On the Role of Temperature Sampling in Test-Time Scaling2510.02611v1 · Yuheng Wu, Azalia Mirhoseini, Thierry Tambe2025 · 1 citationarXiv
Training of physical neural networks.40903603 · Momeni, Ali, Rahmani, Babak, Scellier, Benjamin et al.2025 · 79 citationsNature. 2025;645(8079):53-61
Cartridges: Lightweight and general-purpose long context representations via self-study2506.06266v3 · Sabri Eyuboglu, Ryan Ehrlich, Simran Arora et al.2025 · 0 citationsarXiv
Archon: An Architecture Search Framework for Inference-Time Techniques2409.15254v6 · Jon Saad-Falcon, Adrian Gamarra Lafuente, Shlok Natarajan et al.2024 · 0 citationsarXiv
Exploring Diffusion Transformer Designs via Grafting2506.05340v2 · Keshigeyan Chandrasegaran, Michael Poli, Daniel Y. Fu et al.2025 · 0 citationsarXiv
Synthetic Data Generation & Multi-Step RL for Reasoning & Tool Use2504.04736v2 · Anna Goldie, Azalia Mirhoseini, Hao Zhou et al.2025 · 0 citationsarXiv
Think, Prune, Train, Improve: Scaling Reasoning without Scaling Models2504.18116v1 · Caia Costello, Simon Guo, Anna Goldie et al.2025 · 0 citationsarXiv
Reasoning-SQL: Reinforcement Learning with SQL Tailored Partial Rewards for Reasoning-Enhanced Text-to-SQL2503.23157v2 · Mohammadreza Pourreza, Shayan Talaei, Ruoxi Sun et al.2025 · 1 citationarXiv
How Do Large Language Monkeys Get Their Power (Laws)?2502.17578v1 · Rylan Schaeffer, Joshua Kazdan, John Hughes et al.2025 · 0 citationsarXiv
KernelBench: Can LLMs Write Efficient GPU Kernels?2502.10517v1 · Anne Ouyang, Simon Guo, Simran Arora et al.2025 · 1 citationarXiv
CodeMonkeys: Scaling Test-Time Compute for Software Engineering2501.14723v2 · Ryan Ehrlich, Bradley Brown, Jordan Juravsky et al.2025 · 0 citationsarXiv
Large Language Monkeys: Scaling Inference Compute with Repeated Sampling2407.21787v3 · Bradley Brown, Jordan Juravsky, Ryan Ehrlich et al.2024 · 12 citationsarXiv
Addendum: A graph placement methodology for fast chip design.39327495 · Goldie, Anna, Mirhoseini, Azalia, Yazgan, Mustafa et al.2024 · 11 citationsNature. 2024;634(8034):E10-E11
CHESS: Contextual Harnessing for Efficient SQL Synthesis2405.16755v3 · Shayan Talaei, Mohammadreza Pourreza, Yu-Chen Chang et al.2024 · 9 citationsarXiv
That Chip Has Sailed: A Critique of Unfounded Skepticism Around AI for Chip Design2411.10053v1 · Anna Goldie, Azalia Mirhoseini, Jeff Dean2024 · 0 citationsarXiv
CATS: Contextually-Aware Thresholding for Sparsity in Large Language Models2404.08763v4 · Donghyun Lee, Je-Yong Lee, Genghan Zhang et al.2024 · 0 citationsarXiv
Author Correction: A graph placement methodology for fast chip design.35361999 · Mirhoseini, Azalia, Goldie, Anna, Yazgan, Mustafa et al.2024 · 5 citationsNature. 2022;604(7906):E24
Training of Physical Neural Networks2406.03372v1 · Ali Momeni, Babak Rahmani, Benjamin Scellier et al.2024 · 79 citationsarXiv
Hydragen: High-Throughput LLM Inference with Shared Prefixes2402.05099v2 · Jordan Juravsky, Bradley Brown, Ryan Ehrlich et al.2024 · 1 citationarXiv
Specific versus General Principles for Constitutional AI2310.13798v1 · Sandipan Kundu, Yuntao Bai, Saurav Kadavath et al.2023 · 7 citationsarXiv
FLAME: Taming Backdoors in Federated Learning (Extended Version 1)2101.02281v5 · Thien Duc Nguyen, Phillip Rieger, Huili Chen et al.2021 · 4 citationsarXiv
Embroid: Unsupervised Prediction Smoothing Can Improve Few-Shot Classification2307.11031v1 · Neel Guha, Mayee F. Chen, Kush Bhatia et al.2023 · 4 citationsarXiv
The Capacity for Moral Self-Correction in Large Language Models2302.07459v2 · Deep Ganguli, Amanda Askell, Nicholas Schiefer et al.2023 · 53 citationsarXiv
Constitutional AI: Harmlessness from AI Feedback2212.08073v1 · Yuntao Bai, Saurav Kadavath, Sandipan Kundu et al.2022 · 322 citationsarXivon Valency
Measuring Progress on Scalable Oversight for Large Language Models2211.03540v2 · Samuel R. Bowman, Jeeyoon Hyun, Ethan Perez et al.2022 · 34 citationsarXiv
A Full-Stack Search Technique for Domain Optimized Deep Learning Accelerators2105.12842v3 · Dan Zhang, Safeen Huda, Ebrahim Songhori et al.2021 · 50 citationsarXiv
Representing Long-Range Context for Graph Neural Networks with Global Attention2201.08821v1 · Zhanghao Wu, Paras Jain, Matthew A. Wright et al.2022 · 139 citationsarXiv
A Transferable Approach for Partitioning Machine Learning Models on Multi-Chip-Modules2112.04041v1 · Xinfeng Xie, Prakash Prabhu, Ulysse Beaugnon et al.2021 · 3 citationsarXiv
Delving into Macro Placement with Reinforcement Learning2109.02587v1 · Zixuan Jiang, Ebrahim Songhori, Shen Wang et al.2021 · 12 citationsarXiv
A graph placement methodology for fast chip design.34108699 · Mirhoseini, Azalia, Goldie, Anna, Yazgan, Mustafa et al.2021 · 612 citationsNature. 2021;594(7862):207-212
Transferable Graph Optimizers for ML Compilers2010.12438v2 · Yanqi Zhou, Sudip Roy, Amirali Abdolrashidi et al.2020 · 0 citationsNeurIPS 2020
Chip Placement with Deep Reinforcement Learning2004.10746v1 · Azalia Mirhoseini, Anna Goldie, Mustafa Yazgan et al.2020 · 151 citationsarXiv
Placement Optimization with Deep Reinforcement Learning2003.08445v1 · Anna Goldie, Azalia Mirhoseini2020 · 38 citationsarXiv
RankMap: A Framework for Distributed Learning From Dense Data Sets.28534788 · Mirhoseini, Azalia, Dyer, Eva L, Songhori, Ebrahim M et al.2019 · 1 citationIEEE transactions on neural networks and learning systems. 2018;29(7):2717-2730
Generalized Clustering by Learning to Optimize Expected Normalized Cuts1910.07623v1 · Azade Nazi, Will Hang, Anna Goldie et al.2019 · 0 citationsarXiv
Career total: 121 works. 58 are in this corpus.Showing the 50 most recent.
Profile built from the corpus for this byline.
Author records are still filling in while the Hub is in alpha. If this is your page, you'll be able to claim it soon. Spot a mistake? Tell us.