Research Scientist at ByteDance Seed (Singapore), working on LLM agents and reinforcement learning. Previously at Salesforce Research. Ph.D. in Computer Science from SUTD.
- 🧮 Core contributor to Seed-Prover — Lean 4 theorem-proving agents trained with large-scale RL
- 🤖 Interested in agentic RL: long-horizon credit assignment, reward design, and learning from experience
- Seed-Prover 1.5: Mastering Undergraduate-Level Theorem Proving via Learning from Experience (2025) Agentic RL + test-time scaling for formal proving — solves 88% of PutnamBench, 80% of Fate-H (graduate), 33% of Fate-X (PhD-level), and 11/12 Putnam 2025 problems within 9 hours.
- Seed-Prover: Deep and Broad Reasoning for Automated Theorem Proving (2025) Lemma-style whole-proof reasoning with deep & broad test-time inference for IMO-level formal math.
🌐 allanj.github.io · 🎓 Google Scholar
🤖 This profile README was summarized and created by Claude