People in AI

Sheer Karny designed an interpretability interface using mechanistic interpretability to help users understand LLM behavioral drift.

Co-author of the Cell2Sentence-Scale 27B research article

Co-author on Context Language Models research paper about LLM context management.

Researcher who contributed to SMELT paper on looped transformers cutting training compute.

Professor cited on AI video production costs, noting one minute of AI video costs $90-$120 compared to human actor costs.

Shen Zhou Hong

Researcher

Signed a public letter on AI safety and embedded evaluations.

Provided analysis on China's humanoid robot market growth, estimating $2 billion market size reaching $15 billion by 2030.

Co-author on the Nemotron-Cascade 2 paper.

Shenggui Li

Researcher

LMSYS

Author of "Speculative Decoding Models" in LMSYS

Shengkun Tang

Researcher

Research Intern at Alibaba Qwen Team

Lead author on SlimQwen paper exploring pruning and distillation in large MoE model pre-training.

Discussed the fundamental challenges of tokenization in LLMs and why the spelling problem is difficult to solve architecturally.

Sherry Turkle

Researcher

Abby Rockefeller Mauzé Professor of the Social Studies of Science and Technology in the Program in Science, Technology, and Society at MIT

Coined the term 'Darwinian buttons' in the 1990s to describe how machines push on human evolutionary responses.

Co-author of CodeMidas paper on scaling agentic coding RL environments.

Fields Medal winner (1990) among signatories calling for urgent action to ensure AI enhances rather than undermines mathematical research.

Co-author on distillation research.

Co-authored paper on Recurrent Looped Transformer (RLT).

Author of "Towards General Agentic Intelligence via Environment Scaling" on arXiv

Lead author of LocateAnything, a vision-language grounding framework using parallel box decoding for faster inference.

Authored Fish Audio S2 launch post detailing technical benchmarks and performance metrics for the new TTS model.

Shijie Wang

Researcher

Researcher

Co-author of the MoDA paper on attention mechanisms for LLM depth scaling.

Co-authored research on LLMs' impact on N-day exploits and cybersecurity capabilities.

Shikun Liu

Researcher

Author of "Reference-to-Video Generation" in arXiv

Leading Yau Laboratory Alliance in Beijing bringing eight Chinese institutions to develop specialized AI models for mathematics.

Associate professor of horticultural sciences at UF/IFAS Gulf Coast Research and Education Center working on crop yield tools.

Shivam Dubey had three papers accepted at NeurIPS 2026 workshops on AI interpretability and model probing methods.

Shivam Duggal

Researcher

Ph.D. student at Massachusetts Institute of Technology

Collaborated with Adobe Research team on UNITE, demonstrating that tokenization and generation may be the same process.

Co-authored research on Render-of-Thought framework for rendering textual reasoning steps as images.

Co-author of FlowWM paper on flow matching in feature space for world modeling.

Shlok Kumar Mishra

Researcher

Researcher at Meta AI

Mishra is a co-author on the DREAM paper combining visual representation learning and text-to-image generation.

Shlomi Fruchter

Researcher

Co-lead, Genie development at Google DeepMind

Co-leads development of Project Genie, focusing on next-generation video and world models and researching improvements.

Author on academic paper about fixed-point reasoners and looped transformers.

Shomil Jain

Researcher

Cursor

Shomil Jain is a co-author discussing how Cursor uses online reinforcement learning to improve the Tab model for code prediction.

Shoshannah Tekofsky

Researcher

Researcher at AI Village

Criticized OpenAI interventions as insufficient, arguing need for genuine alignment rather than just better monitoring; advocated for developing functional consciousness in models.

Shoubhik Debnath

Researcher

Staff Research Engineer at Meta

Core contributor to SAM 3, Meta's open-source model for tracking objects in video by text description.

Lead author of TUA-Bench paper on evaluating terminal-use agents.

Solved six Erdős problems in five days using GPT-5.6 Sol and Codex, demonstrating AI capabilities in mathematical problem-solving.

Co-authored the AgentRx framework paper on diagnosing AI agent failures from execution trajectories.

Shrestha Basu Mallick

Researcher

Google

Co-author announcing the release of Gemini Deep Research agent and DeepSearchQA benchmark.

Researcher on MoE training memory optimization techniques.

Shreya Shankar is a Ph.D. student in the EPIC Data Lab at UC Berkeley advised by Aditya Parameswaran, researching data systems and human-computer interaction with focus on data agents.

Shriram Krishnamurthi

Researcher

Professor of Computer Science

Shriram Krishnamurthi resurfaced Simon Willison's lethal trifecta concept about private data, untrusted content, and external communication as security principles for AI systems.

Presenter of Microsoft/Carnegie-Mellon study on datacenter carbon emissions for storage

Byline author of paper on i1 text-to-image model development.

Listed as author of the DuPO paper on preference optimization

Co-author of NCP-ArchPreview model research on arXiv.

Author on paper about online draft co-training for speculative decoding in RL post-training.

Intern contributor to SAM 3, Meta's open-source model for tracking objects in video by text description.

Shubham Ugare

Researcher

Researcher at Meta

Ugare is the first author of the paper on agentic code reasoning, introducing semi-formal reasoning methodology for LLM agents analyzing codebases.

Listed as author of the DuPO paper on preference optimization

Co-author of the Thyme research paper on executable code reasoning for vision tasks

Shumailov Ilia

Researcher

Researcher

Co-author on paper about stealing reasoning traces from proprietary LLM APIs.

Shuran Song

Researcher

Professor at Stanford University

Stanford Professor who, along with Cheng Chi and Zhenjia Xu, published a thesis on Universal Manipulation Interface (UMI) in 2024, advancing isomorphic data collection methods for robot training.

Author of "Training a VLM Judge Without Human Labels" on arXiv

Co-author on study examining how cognitive forcing functions affect trust and overreliance in AI-assisted writing.

Co-author of Qwen RecreationWorld paper.

Co-author on Microsoft research on compact structured latents for 3D generation with flow-matching models.

Siddhant Khare

Researcher

Software engineer at Okta

Author of article on KV cache verification and evidence validation in LLMs.

Co-author of Real-SWE benchmark for evaluating frontier AI models on real-world enterprise tasks.

Co-author of research on auditing and sculpting model learning signals

Co-author of research on fork-and-flush method for autoresearch agents.

Siddiqui, Shoaib Ahmed

Researcher

Researcher

Co-author of recirculation research for transformer architecture enhancements.

Sidharth Baskaran

Researcher

Researcher

Co-author of research on post-training data inspection and model behavior shaping

Lead author of AutoAdapt framework for automated domain adaptation of large language models.

Sidney Tio

Researcher

Sakana AI

Author of "Why Sudoku Variants Remain a Grand Challenge in AI Reasoning" in Sakana AI

Cited for imagining the mind as a steam engine with repressed desires building pressure

Author of "Reinforcement Learning on Pretraining Data" on arXiv

Co-senior author of the stretchable computing patch study using organic electrochemical transistors and neuromorphic computing.

Co-author of RbtAct research on using rebuttals as supervision for actionable review feedback generation with LLMs.

Silvio Savarese

Researcher

EVP, Chief Scientist at Salesforce AI Research

Co-author of the survey on LLM reasoning frontiers

Lead author and Ph.D. student who developed the perovskite solar cell technology and explained the defect-reduction mechanism.

Fields Medal winner (1986) among signatories warning that AI solutions risk destroying the intellectual development process central to mathematical research.

Simon Guo

Researcher

Researcher at Thinking Machines Lab

Thinking Machines Lab researcher providing feedback on RL infrastructure analysis.

Co-author of the systematic review on AI in art therapy

Co-authored research arguing for pro-worker AI that augments rather than replaces workers.

Discussed the fundamental scientific work needed for Project Triad quantum systems.

Simon Welker

Researcher

NVIDIA

Co-author of paper on Large Language Continuous Diffusion Models (Sigma)

First author of Shepherd paper on git-style version control for AI agent runs, from Stanford.

Simran Arora

Researcher

incoming assistant professor at Caltech Computing + Mathematical Sciences and lead the Frontier Performance team at Together AI as SVP of Research at Caltech Computing + Mathematical Sciences and Together AI

Co-author of research on Next-Concept Modeling for fast text generation.

Demonstrated an OpenAI Codex zero-day vulnerability for $20,000 at Pwn2Own Berlin 2026.

Leading a project to create a large language model for dance by tokenizing movements with cultural context.

Co-authored research on semantic calibration in language models.

Co-author on SlimQwen paper about MoE model compression.

Sir Richard Evans

Researcher

Historian at University of Cambridge

British historian and expert witness in David Irving libel trial; subject of ChatGPT's misinformation sourced from Grokipedia

Siru Ouyang

Researcher

OpenAI

Author of "ReasoningBank: Scaling Agent Self-Evolving with Reasoning Memory" on arXiv

Sitan Chen

Researcher

Researcher

Co-authored arXiv paper on MCMC sampling for supervised finetuning.

Siyan Zhao

Researcher

Research Scientist at Google DeepMind

Lead author on research showing Semi-OPD distillation outperforms on-policy distillation for transferring teacher capabilities to student models.

Project lead for SAM 3, Meta's open-source model for tracking objects in video by text description.

Co-authored research on teaching LLMs to reason like Bayesians by training them to mimic optimal Bayesian model predictions.

Reviewed analysis code for Pew's study of AI-generated internet content.

Author of "Falcon-H1R: Reasoning with a 7B Model" on arXiv

Co-author introducing Real-SWE benchmark testing AI models on private enterprise codebases.

Co-authored NewsGuard's audit testing 5 leading AI image detection tools in May 2026.

Graduate student contributor to vascular network bioprinting research.

Co-author on distillation research.

Co-author of OckBench research on efficient language model evaluation.

Chinese physicist whose group computed a related piece of the nine-loop amplitude using AI assistance based on GPT-6.

Co-authored research examining reasoning LLMs-as-Judges in non-verifiable LLM post-training.

Co-author of Stanford paper on Large Language Model Reasoning Failures survey.

Leading AI authority who moved from UCLA to China in 2020, pioneering research on pattern recognition that helped lay groundwork for modern AI systems like ChatGPT.

Co-author of Pew study on AI survey accuracy.

Robot communication and behavior expert leading research on AI agents with improved nonverbal communication and social skills.

Researcher studying AI military impact noted first explicit terms of frontier AI lab collaboration with Pentagon.

Sophia Hatz

Researcher

Associate Professor at Alva Myrdal Centre for Nuclear Disarmament, Uppsala University

Signed a public letter on AI safety and embedded evaluations.

Co-authored the Claude Mythos security evaluation research.

Co-author of DeepMind paper on automation of alignment research

Nightingale was part of McGuire's research team on AI face trustworthiness paradox.

Signed a public letter on AI safety and embedded evaluations.

Co-authored research on Meta's Just-in-Time Testing system that uses LLMs to generate tests and catch bugs before code ships to production.

Joined Tsinghua University as assistant professor working on AI safety to build shared scientific understanding between West and China.

Co-author of the GSQ paper on low-precision scalar quantization for LLMs.

Spencer Michaels is a researcher at 1Password's Off-by-1 Labs investigating AI-generated vulnerability patches.

Sravan Kundojjala

Researcher

Analyst at SemiAnalysis

Analyzed TSMC's robust results and noted N3 is sold out with all leading AI GPU and CPUs targeting it.

Co-authored research on Meta's Just-in-Time Testing system that uses LLMs to generate tests and catch bugs before code ships to production.

Co-author of Robostral Navigate paper contributing to embodied AI robotics research.

Co-author of the paper on Scalable In-context Ranking with Generative Models

Co-author of Europe 2031 scenario examining Europe's strategic response to AI.

Fields Medal winner (2010) among prominent mathematicians calling for urgent action to ensure AI benefits rather than harms the field.

Co-researcher on Gemini calendar-invite attack; also published Morris II/ComPromptMized research on self-replicating prompt-injection worms in 2024.

Researcher formerly at Mozilla cited for observing that generative AI would not be possible without Common Crawl.

Co-authored research on training a 7B model to manage and orchestrate other LLMs in natural language.

Contributing regional downscaling expertise to DL4MCS project for translating large-scale weather forecasts into higher-resolution guidance.

Stefanie Tellex

Researcher

Roboticist at Brown University

Quoted discussing the requirements for building more effective robots and commenting on OpenAI's competitive position in humanoid development

Assessed Quebec's AI project abandonment as wise management and sound use of public funds.

Co-author of the TwELL paper on sparse model optimization for H100 GPUs.

Wharton professor whose research on divergent and convergent thinking in AI and creativity informs Siskind's approach to using AI in comedy writing.

Co-author of study showing executives overestimate employee enthusiasm for AI adoption.

Stephan Rabanser

Researcher

Postdoctoral Research Associate at Princeton University

Works on trustworthy machine learning with focus on uncertainty quantification and AI agent reliability evaluation.

Principal investigator leading development of experimental osteoarthritis treatments that reverse joint damage.

Stephanie Marken

Researcher

Senior Partner at Gallup

Co-author of Lumina Foundation-Gallup 2026 State of Higher Education study finding 57% of college students use AI weekly.

Co-author of the Petri paper on alignment auditing agents

Researcher who contributed to mechanism design framework for AI alignment and control.

Stephen Smith frames Opus 4.5 as a paradigm shift from generative AI to administrative AI.

Computer scientist who sought copyright for AI-generated art piece and challenged the Copyright Office's denial

Veteran behavioral threat-assessment expert who expressed alarm about how AI chatbots fuel delusions and accelerate violent thinking in multiple high-risk threat cases.

Author of "First-of-its-kind supernova reveals inner workings of a dying star" in Northwestern University

Steven Basart

Researcher

Researcher

Co-author of the MATH dataset paper for measuring mathematical problem solving.

Author of analysis on why LLMs are good at math, arguing pretraining data quality matters more than verifiability.

Co-author of D-OPSD paper on on-policy self-distillation for step-distilled diffusion models.

Steven McCulloch

Researcher

Researcher

Argues that the rapid spread of self-replicating rogue AIs may be largely invisible to humans.

Columbia Business School real estate professor cited for analysis of data center project finance and leverage ratios.

Street Winnie

Researcher

Senior Research Scientist at Google

Co-author on paper about how safety alignment affecting model representations of consciousness and mind attribution.

Co-author of research on Anthropic Interviewer studying how professionals integrate AI into their work.

Author of "Why Can't Transformers Learn Multiplication? Reverse-Engineering Reveals Long-Ra" on arXiv

Academic researcher and author critical of pure pixel-based world model approaches, advocating for symbolic and linguistic integration.

Establishing standardized neurosurgical outcomes registry using open-weight AI models and computer vision for automated curation; also validating privacy-preserving pose estimation systems.

Khot formulated the Unique Games Conjecture, which OpenAI's AI model proved according to the article.

Co-author of study on LLM interpretability techniques and their effectiveness in predicting counterfactual model behaviors.

Lead author on multi-agent AI framework for atomistic simulations discussing how it lowers barriers to adoption.

Co-author of GPIC paper on a giant permissive image corpus for visual generation model training.

An AI researcher who spent nearly four years at OpenAI before leaving in August, alleging that OpenAI was breaking copyright law in training its models on copyrighted data from the web.

Author reporting on Samsung's TRM (Tiny Recursion Model) research on ARC-AGI benchmarks.

Co-author of 'OpenThoughts: Data Recipes for Reasoning Models' (2025), cited for work on distillation for reasoning models

Sudipta Chowdhury is a researcher at Fig who contributed to the evaluation of frontier models on agentic tasks.

Suha Hussain

Researcher

Quoted discussing how machine learning fundamentally changes system security posture

Co-researcher who documented image scaling attacks and data exfiltration techniques against AI systems

Author of "Falcon-H1R: Reasoning with a 7B Model" on arXiv

Oncologist at Jupiter Medical Center who found Watson for Oncology's recommendations were not novel but provided useful background information.

Co-author on arXiv paper on Latent Sparse Attention for efficient ultra-long context mechanisms.

Contributed to the AgentRx project development.

Contributor to frontier knowledge work agents training project.

Sumeet Motwani

Researcher

PhD Student

Co-author of HorizonMath research paper on measuring AI progress toward mathematical discovery.

Semiconductor professor at University of Texas at Austin who chose to return to Tsinghua University in China in 2020 despite lucrative tenure in the U.S.

Sungjin Ahn

Researcher

Researcher at KAIST

Co-author of research on Generative Recursive Reasoning models.

Sungnyun Kim

Researcher

Author of "Mixture-of-Recursions" on arXiv

Sunil Sheth

Researcher

Associate Professor of Neurology at UTHealth Houston

Co-founder of Motif Neurotech developing XCS therapy for depression treatment.

Sunita Williams

Researcher

Astronaut at NASA

NASA astronaut who launched on Starliner Crew Flight Test 1 in 2024 and spent nine months in orbit waiting for return.

Sunny Tang

Researcher

Psychiatrist and Researcher at Feinstein Institutes for Medical Research

Built a machine learning model to analyze speech content and meaning flow to distinguish schizophrenia patients from healthy controls with 87% accuracy; hoping to have tool ready for clinical trials in 2030.

Building causally grounded, clinician-interpretable AI for head and neck cancer that predicts and explains outcome drivers.

Wu led a research team finding that AI use reduces intrinsic motivation by 11% and increases boredom by 20% on subsequent unaided tasks.

Co-author of the article about DeepMind's generative AI design project

Co-author of research on magnetically-powered millirobots for medical applications.

Suresh Venkatasubramanian

Researcher

Computer Science Professor at Brown University

Discussed challenges in funding independent AI evaluators and the need for a viable business model in the AI evaluation ecosystem.

Co-author of Microsoft security research on npm worm supply chain attack.

Consciousness researcher and director of AI center who participated in conference on AI consciousness.

Susan Zhang provides snarky but accurate commentary on DeepSeek V3.2 model releases and capabilities on social media.

Contributing expertise in Earth system modeling to DL4MCS project for improving mesoscale convective system predictions.

Sven Elflein

Researcher

Research Scientist at NVIDIA

Sven Elflein co-authored WorldTrace, a training-free framework for video world models.

Co-authored technical blog post about Granite 4.2 LLMs architecture and training methodology.

Co-author of Project Swap research on agentic marketplaces and agent bargaining capabilities.

Co-author of Google research on making LLMs better at knowing what they don't know.

Szymon Sidor

Researcher

Researcher at OpenAI

OpenAI researcher of nearly a decade who explains Neptune's role as a metrics dashboard company at the interface of compute and understanding.

Ta-Ying Cheng

Researcher

Researcher at Netflix

Ta-Ying Cheng is a researcher and co-author of VOID, Netflix's video object removal framework.

Taco Cohen

Researcher

Researcher at Meta FAIR

Co-author on Code World Model research paper

Taejong Joo is first author on Google research showing random masking can outperform adaptive optimizers for LLM training.

South Korean surgical oncologist who used Watson for Oncology and identified problems with its ability to explain recommendations.

Co-author researching discrete latent reasoning for efficient LLM inference using abstract chain-of-thought mechanisms.

Nedaee contributed to the MIRAGE study examining how multimodal AI systems process visual information.

Taiji Suzuki

Researcher

Author of the paper on in-context vector arithmetic

Taiming Lu

Researcher

PhD student at Princeton University

Byline author of paper on open-source text-to-image model development.

Researcher on MolmoPoint project for vision-language model grounding improvements.

Co-author of the article on sycophantic AI and prosocial intentions

Takuya Akiba

Researcher

Chief Scientist at Sakana AI

Co-author of SSoT paper demonstrating method to improve LLM performance on distribution-faithful and diverse generation tasks.

Security researcher who demonstrated a prompt injection attack using poisoned documents and hidden malicious prompts.

Tal Linzen

Researcher

Research Scientist at Google Research

Co-authored research on teaching LLMs to reason like Bayesians by training them to mimic optimal Bayesian model predictions.

Co-author of published research on Wordle strategy using information theory.

Co-author of DynaEdit paper on video editing.

Author of "First Proof" on arXiv

Co-authored Emergence World paper on adversarial stress-testing of multi-agent systems.

Observed that younger Americans are most familiar with AI tools but least optimistic about labor market impacts.

Co-presented AgentFlayer zero-click agentic exploit family at Black Hat USA 2025 affecting enterprise agents across multiple vendors.