I. Opening

Lang Xiong

Stanford CS + Math · National Master

AI researcher, chess master, and community leader. I'm interested in world models and agent evaluations for AI safety.

Scroll to play ↓

II. Middlegame

Experience

• 2026
  Terminal-Bench Task HardnessResearch project · separating genuine from fake hardness · NeurIPS Workshop
• 2024–
  Research with Berkeley, Stanford & MITLLM evaluation awareness, sarcasm, manipulation benchmarks, medical imaging
• 2025–
  Research Intern, EDIT LabDartmouth Hitchcock · leading bone marrow cell segmentation (HoVer-Net, GAT)
• 2025–
  Research Intern, CASE LabUniversity of Maryland · CogniPair, a Global Workspace agent architecture
• 2025
  Market Analyst InternReinsurance Group of America · scouted 20+ early-stage AI startups

III. Tactics

Projects

• 2025
  AgoraHackathon · nonpartisan political literacy app with AI debates
  +
• 2025
  24 Card GamePersonal project · multiplayer web game with online lobbies
  +
• 2024–25
  Ultrasound scoliosis detectionCNN + quantitative ultrasound · provisional patent
  +
• 2023–26
  Sustainable materialsAlgal bioplastic → ML-optimized concrete · 2× ISEF finalist
  +
• 2025
  Autonomous underwater vehicleRaspberry Pi 5, custom PCBs · 6 directions of movement
  +
• 2022–26
  FTC Robotics, captainJava path planning, PID tuning, AprilTag vision
  +
• 2022–26
  American Rocketry ChallengeClub president · 20+ test flights
  +
• 2022–26
  CyberPatriot, team lead2× Platinum, 2× Gold division
  +

Tap a project for details and photos.

IV. Notation

Notable Publications

• 2025
  Probe-Rewrite-Evaluate: A Workflow for Reliable Benchmarks and Quantifying Evaluation AwarenessFirst author · NeurIPS Workshop
• 2025
  Sarc7: Evaluating Sarcasm Detection and Generation with Seven Types and Emotion-Informed TechniquesFirst author · COLM & EMNLP Workshops
• 2026
  What Makes a Terminal-Bench Task Hard? Separating Genuine Hardness from Fake-Hardness on an Adjudicated Agentic CorpusNeurIPS Workshop

All papers on Google Scholar · 23 citations →

V. Endgame

Impact

• 2022–
  Checkmate4Change ↗Founder · chess nonprofit, 9 chapters, $20k raised
• 2022–
  BioThrive ↗Founder · 250+ volunteers, 21 restoration projects
• 2021–
  Eagle ScoutSenior Patrol Leader, Troop 869

VI. Titles

Honors

• 2026
  Coca-Cola Scholar1 of 150 from 107,000+
• 2026
  LinkedIn Possibilities in Tech Scholar1 of 25
• 2026
  UToronto Pearson ScholarFull ride · 37 recipients worldwide
• 2026
  Jane Street ML SummitInvited participant · 1 of ~40 students
• 2025–26
  2× ISEF FinalistUnder 1% acceptance rate
• 2025–26
  2× AIME QualifierAmerican Invitational Mathematics Examination
• 2026
  USAPhO SemifinalistUS Physics Olympiad
• 2025
  Rensselaer Medal · RIT Math and Science Award · Bausch & Lomb AwardRensselaer Polytechnic Institute · Rochester Institute of Technology · University of Rochester
• 2023
  US Chess National Master2200 USCF rating · peak #33 in age group

VII. Off the board

Interests

Chess · Go · Bridge · Guandan 掼蛋 · Guitar · Cooking · Badminton · C-dramas · Skiing · Gym · Pickleball · Content creation

• Create
  WeChat Channel 微信视频号My Chinese–English channel, running since I was 8

• Eat
  Food recs in China

  1. Fei Da Chu 费大厨 · Hunan pepper-fried pork
  2. Chaoshan beef hot pot 潮汕牛肉火锅
  3. Yunnan straw-hat stone pot fish 草帽石锅鱼 · steamed under a straw hat
  4. Lao Wan Hui 老碗会 · Shaanxi noodles

• Book
  The Three-Body ProblemLiu Cixin · favorite book
• Watch
  Top C-drama picks

  1. Spy Game 特工任务
  2. Storm Eye 暴风眼
  3. Forging Justice 重器
  4. Three-Body 三体
  5. Amidst a Snowstorm of Love 在暴雪时分
  6. The Winter Solstice 冬至

VIII. Your move

Get in touch

Open to research collaborations, internships, and good chess games.

White to move, mate in 1. Click a piece, then a square.

© 2026 Lang Xiong · GG