Hi, I’m Soren Dunn. I’m a Founding Engineer at Intology AI. I was previously a master’s student at the University of Illinois at Urbana-Champaign, supervised by Professor Lingming Zhang. I received my bachelor’s, triple majoring in Data Science, Statistics, and Chemistry from the University of Chicago in 2023.
I am interested in pushing the frontiers of speeding up technological progress and scientific discovery using AI. Along the way, I have contributed to frameworks for evaluating LLM’s SWE capabilities and work in automating scientific discovery.
Contact
- Email: soren@intology.ai
- GitHub: @sorendunn
- Twitter: @soren_dunn_
- CV: Resume
News
- May 27th, 2025: Intology’s artifical scientist Zochi published in ACL 2025, “Tempest: Automatic Multi-Turn Jailbreaking of Large Language Models with Tree Search” - the first fully autonomous publication in a top scientific venue by an AI system.
- March 14th, 2025: Along with Intology AI, I debuted Zochi, an artificial scientist which produced papers averaging 7.67/10 on NeurIPS guidelines, 3 of which were accepted to ICLR 2025 workshops.
- February 26th, 2025: Agentless Lite doubles state-of-the-art on SWE-bench Multimodal from 12.19% to 25.34% for 1/4th of the cost without even requiring a runtime environment!
- February 14th, 2025: I released Agentless Lite - a generalized, lightweight adaptation of the Agentless scaffold which is competitive with SOTA agents while only requiring sampling from a single RAG prompt.
- January 31st, 2025: Agentless used by both DeepSeek and OpenAI to evaluate their new reasoning models (r1 and o3-mini) on SWE-bench
- Dec 2nd, 2024: Integrated Agentless with Claude 3.5 Sonnet to achieve 40.7% solve rate on SWE-bench lite and 50.8% solve rate on SWE-bench verified
- Oct 28th, 2024: Released OpenAutoCoder-Agentless 1.5 which increases Agentless performance from 27.3% to 32.00% on SWE-bench lite
- September 26, 2024: MedCalc-Bench was accepted as an oral presentation for the NeurIPS 2024 Datasets and Benchmark Track
- September 12, 2024: Agentess was used by OpenAI as their scaffold of choice for evalauting gpt-4o, o1-mini, and o1-preview’s model autonomy as part of their preparedness framework
- July 1st, 2024: Released OpenAutoCoder-Agentless 1.0! Agentless currently is the best open-source approach on SWE-bench lite with 82 fixes (27.3%) and costing on average $0.34 per issue.
Education
- Data Science (B.S.) with Honors, Statistics (B.A.), Chemistry (B.A.), University of Chicago October 2019 - June 2023
- Masters of Statistics, University of Illinois at Urbana-Champaign, August 2023 - June 2025
Experience
- Intology AI, Founding Engineer, San Francisco, CA, January 2025 - Present
- Developed experimentation, ideation, paper-writing capabilities, and user interface for Zochi, the first AI system to publish a paper in a top scientific venue (ACL 2025) with reviews placing it in the top 8.2% of all ACL submissions as well as publishing three papers in ICLR 2025 workshops
- Configured Zochi to run on MLE-bench, reaching SOTA on a canonical subset of MLE-Bench achieving medals in 50% of Kaggle competitions it was tested on
Publications
[FSE 2025 Distinguished Paper] Chunqiu Steven Xia*, Yinlin Deng*, Soren Dunn, Lingming Zhang. Agentless: Demystifying LLM-based Software Engineering Agents, July, 2024 [paper] [code]
[NeurIPS 2024 Datasets and Benchmark Track Oral] Nikhil Khandekar*, Qiao Jin*, Guangzhi Xiong*, Soren Dunn, Serina S Applebaum, Zain Anwar, Maame Sarfo-Gyamfi, Conrad W Safranek, Abid Anwar, Andrew Jiaxing Zhang, Aidan Gilson, Maxwell B Singer, Amisha D Dave, R. Andrew Taylor, Aidong Zhang, Qingyu Chen, Zhiyong Lu. MedCalc-Bench: Evaluating Large Language Models for Medical Calculations, June, 2024 [paper] [code]
[Journal of the American Chemical Society] Xuanyu Feng, Yang Song, Justin Chen, Ziwan Xu, Soren Dunn, Wenbin Lin. Rational Construction of an Artificial Binuclear Copper Monooxygenase in a Metal–Organic Framework, January, 2021 [paper]